1269 lines
53 KiB
C#
1269 lines
53 KiB
C#
using iTextSharp.text.pdf.parser.clipper;
|
||
using MathNet.Numerics.Interpolation;
|
||
using System;
|
||
using System.Collections.Generic;
|
||
using System.Linq;
|
||
using System.Text;
|
||
|
||
namespace ConductivityApp.GBStandard
|
||
{
|
||
|
||
|
||
/// <summary>
|
||
/// GB/T 32064-2015 标准计算器（严格遵循国标）
|
||
/// 瞬态平面热源测试法计算导热系数和热扩散系数
|
||
/// </summary>
|
||
public class GB32064Calculator
|
||
{
|
||
#region 结构定义
|
||
public struct BridgeConfig
|
||
{
|
||
public double Rs { get; set; } // 串联电阻 (Ω)
|
||
public double RL { get; set; } // 引线电阻 (Ω)
|
||
}
|
||
|
||
public struct ProbeConfig
|
||
{
|
||
public double R0 { get; set; } // 初始电阻 (Ω)
|
||
public double Alpha { get; set; } // 电阻温度系数 (1/K)
|
||
public double RadiusMM { get; set; } // 探头半径 (mm)
|
||
public int CoilCount { get; set; } // 探头环数
|
||
}
|
||
|
||
public struct TestConfig
|
||
{
|
||
public double P0 { get; set; } // 输出功率 (W)
|
||
public double SampleDensity { get; set; } // 样品密度 (kg/m³)
|
||
|
||
public string cmbSampleType { get; set; }
|
||
}
|
||
|
||
public struct CalculationResult
|
||
{
|
||
public bool IsValid { get; set; }
|
||
public double ThermalConductivity { get; set; } // λ - W/(m·K)
|
||
public double ThermalDiffusivity { get; set; } // α - m²/s
|
||
public double SpecificHeatCapacity { get; set; } // Cp - J/(kg·K)
|
||
public double CorrectionTime { get; set; } // tc - s
|
||
public double RSquared { get; set; }
|
||
public string ValidationMessage { get; set; }
|
||
public string CalculationLog { get; set; } // 新增：计算过程日志
|
||
}
|
||
#endregion
|
||
|
||
#region 常量定义（根据国标GB/T 32064-2015）
|
||
private const double TAU_MAX_SQUARED_LOWER = 0.3; // τ_max²的最小值（国标5.3.4要求）
|
||
private const double TAU_MAX_SQUARED_UPPER = 1.0; // τ_max²的最大值（国标5.3.4要求）
|
||
private const double PROBING_DEPTH_RATIO_LOWER = 1.1; // 探测深度/探头半径最小值
|
||
private const double PROBING_DEPTH_RATIO_UPPER = 2.0; // 探测深度/探头半径最大值
|
||
private const double MIN_DATA_COUNT = 100; // 最小数据点数（国标5.3.3.3）
|
||
private const double MIN_TIME_INTERVAL = 0.1; // 最小时间间隔(s)（国标5.3.3.3）
|
||
private double PI_POW_1_5 = Math.Pow(Math.PI, 1.5); // π^(3/2)
|
||
#endregion
|
||
|
||
#region 私有字段
|
||
private readonly BridgeConfig _bridge;
|
||
private readonly ProbeConfig _probe;
|
||
private readonly TestConfig _test;
|
||
private IInterpolation _tauDInterpolation;
|
||
private StringBuilder _logBuilder; // 新增：日志记录器
|
||
#endregion
|
||
|
||
#region 构造函数
|
||
public GB32064Calculator(BridgeConfig bridge, ProbeConfig probe, TestConfig test)
|
||
{
|
||
_bridge = bridge;
|
||
_probe = probe;
|
||
_test = test;
|
||
_logBuilder = new StringBuilder();
|
||
|
||
LogInfo("GB32064Calculator 初始化开始");
|
||
LogInfo($"探头配置: R0={_probe.R0}Ω, α={_probe.Alpha}/K, 半径={_probe.RadiusMM}mm, 环数={_probe.CoilCount}");
|
||
LogInfo($"电桥配置: Rs={_bridge.Rs}Ω, RL={_bridge.RL}Ω");
|
||
LogInfo($"测试配置: P0={_test.P0}W, 密度={_test.SampleDensity}kg/m³");
|
||
|
||
ValidateConfigurations();
|
||
InitializeTauDTable();
|
||
LogInfo("GB32064Calculator 初始化完成");
|
||
}
|
||
#endregion
|
||
|
||
|
||
#region 可配置参数
|
||
private int _maxIterations = 300;
|
||
private double _rSquaredTarget = 0.95;
|
||
private double _convergenceThreshold = 1e-8;
|
||
private int _maxStagnationIterations = 30;
|
||
|
||
/// <summary>
|
||
/// 设置迭代参数
|
||
/// </summary>
|
||
public void SetIterationParameters(int maxIterations = 300, double rSquaredTarget = 0.95,
|
||
double convergenceThreshold = 1e-8, int maxStagnation = 30)
|
||
{
|
||
_maxIterations = maxIterations;
|
||
_rSquaredTarget = rSquaredTarget;
|
||
_convergenceThreshold = convergenceThreshold;
|
||
_maxStagnationIterations = maxStagnation;
|
||
|
||
LogInfo($"设置迭代参数: 最大迭代={_maxIterations}, R²目标={_rSquaredTarget:F3}");
|
||
}
|
||
#endregion
|
||
|
||
|
||
#region 配置验证
|
||
private void ValidateConfigurations()
|
||
{
|
||
LogInfo("开始配置验证");
|
||
var errors = new List<string>();
|
||
|
||
if (_probe.R0 <= 0) errors.Add("探头初始电阻R0必须大于0");
|
||
if (_probe.Alpha <= 0) errors.Add("探头温度系数α必须大于0");
|
||
if (_probe.RadiusMM <= 0) errors.Add("探头半径必须大于0");
|
||
if (_probe.CoilCount < 1) errors.Add("探头环数应大于0");
|
||
|
||
if (_bridge.Rs <= 0) errors.Add("串联电阻Rs必须大于0");
|
||
if (_bridge.RL < 0) errors.Add("引线电阻RL不能为负");
|
||
|
||
if (_test.P0 <= 0) errors.Add("输出功率P0必须大于0");
|
||
if (_test.SampleDensity <= 0) errors.Add("样品密度必须大于0");
|
||
|
||
if (errors.Count > 0)
|
||
{
|
||
LogError($"配置验证失败，发现{errors.Count}个错误");
|
||
throw new ArgumentException($"配置验证失败:\n{string.Join("\n", errors)}");
|
||
}
|
||
else
|
||
{
|
||
LogInfo("配置验证通过");
|
||
}
|
||
}
|
||
#endregion
|
||
|
||
#region 核心计算方法
|
||
|
||
|
||
private double GetDefaultAlphaBySoilType()
|
||
{
|
||
switch (_test.cmbSampleType)
|
||
{
|
||
case "干土":
|
||
return 0.8e-6; // 干土的α更小，修正为0.3e-6
|
||
case "湿土":
|
||
return 2.0e-6; // 湿土典型值
|
||
case "冻土":
|
||
return 2.5e-6; // 冻土典型值
|
||
default:
|
||
return 1.6e-6; // 通用默认值
|
||
}
|
||
}
|
||
|
||
|
||
public CalculationResult Calculate(double[] timeArray, double[] deltaUArray, double currentMA = 120.0)
|
||
{
|
||
LogInfo("=".PadRight(80, '='));
|
||
LogInfo("开始计算导热系数和热扩散系数");
|
||
LogInfo($"输入数据: 时间点数={timeArray?.Length}, 电压点数={deltaUArray?.Length}, 电流={currentMA}mA");
|
||
|
||
try
|
||
{
|
||
// 1. 数据验证
|
||
LogInfo("步骤1: 数据验证");
|
||
if (!ValidateInputData(timeArray, deltaUArray))
|
||
{
|
||
LogError("输入数据验证失败");
|
||
return InvalidResult("输入数据验证失败");
|
||
}
|
||
LogInfo($"数据验证通过: 时间范围[{timeArray.Min():F4}s, {timeArray.Max():F4}s], 共{timeArray.Length}个点");
|
||
|
||
double I0 = currentMA / 1000.0; // mA转A
|
||
LogInfo($"电流转换: {currentMA}mA -> {I0:F4}A");
|
||
|
||
// 2. 计算ΔT(t) - 国标公式(3)
|
||
LogInfo("步骤2: 计算温度增量ΔT(t) - 国标公式(3)");
|
||
double[] deltaT = CalculateTemperatureIncrements(deltaUArray, I0);
|
||
LogInfo($"ΔT计算完成: 平均值={deltaT.Average():E6}K, 最大值={deltaT.Max():E6}K, 最小值={deltaT.Min():E6}K");
|
||
|
||
// 3. 迭代求解热扩散系数α和校正时间tc
|
||
LogInfo("步骤3: 迭代求解热扩散系数α和校正时间tc");
|
||
var (alpha, tc, iterationLog) = IterateThermalDiffusivity(timeArray, deltaT);
|
||
LogInfo(iterationLog); // 记录迭代过程
|
||
LogInfo($"迭代完成: α={alpha:E6} m²/s, tc={tc:F4}s");
|
||
|
||
// 4. 验证τ_max²是否符合国标要求
|
||
LogInfo("步骤4: 验证τ_max²是否符合国标要求");
|
||
double r = _probe.RadiusMM / 1000.0;
|
||
double tmax = timeArray.Max();
|
||
double tauMaxSquared = CalculateTauMaxSquared(tmax, tc, alpha, r);
|
||
LogInfo($"τ_max²计算: tmax={tmax:F2}s, tc={tc:F4}s, α={alpha:E6} m²/s, r={r:F6}m");
|
||
LogInfo($"τ_max²结果: {tauMaxSquared:F4} (要求范围: [{TAU_MAX_SQUARED_LOWER}, {TAU_MAX_SQUARED_UPPER}])");
|
||
|
||
if (tauMaxSquared < TAU_MAX_SQUARED_LOWER || tauMaxSquared > TAU_MAX_SQUARED_UPPER)
|
||
{
|
||
LogError($"τ_max²({tauMaxSquared:F3})不在国标要求范围内");
|
||
return InvalidResult($"τ_max²({tauMaxSquared:F3})不在国标要求范围[{TAU_MAX_SQUARED_LOWER}, {TAU_MAX_SQUARED_UPPER}]内，测试无效");
|
||
}
|
||
LogInfo("τ_max²验证通过");
|
||
|
||
// 5. 计算导热系数λ - 国标公式(4)
|
||
LogInfo("步骤5: 计算导热系数λ - 国标公式(4)");
|
||
double lambda = CalculateThermalConductivity(timeArray, deltaT, alpha, tc);
|
||
LogInfo($"导热系数计算结果: λ={lambda:F6} W/(m·K)");
|
||
|
||
// 6. 计算比热容Cp - Cp = λ / (ρ × α)
|
||
LogInfo("步骤6: 计算比热容Cp");
|
||
double cp = CalculateSpecificHeatCapacity(lambda, alpha);
|
||
LogInfo($"比热容计算结果: Cp={cp:F2} J/(kg·K)");
|
||
|
||
// 7. 验证测试结果有效性
|
||
LogInfo("步骤7: 验证测试结果有效性");
|
||
var validation = ValidateTestResults(timeArray, alpha, tc, tauMaxSquared);
|
||
LogInfo($"测试结果验证: {(validation.IsValid ? "通过" : "未通过")}");
|
||
|
||
// 8. 计算R²
|
||
LogInfo("步骤8: 计算拟合优度R²");
|
||
double rSquared = CalculateRSquared(timeArray, deltaT, alpha, tc, lambda);
|
||
LogInfo($"R²计算结果: {rSquared:F6}");
|
||
|
||
// 生成最终日志
|
||
LogInfo("=".PadRight(80, '='));
|
||
LogInfo("计算完成");
|
||
LogInfo($"最终结果: λ={lambda:F6} W/(m·K), α={alpha:E6} m²/s, Cp={cp:F2} J/(kg·K)");
|
||
LogInfo($"校正参数: tc={tc:F4}s, R²={rSquared:F6}");
|
||
|
||
return new CalculationResult
|
||
{
|
||
IsValid = validation.IsValid,
|
||
ThermalConductivity = lambda,
|
||
ThermalDiffusivity = alpha,
|
||
SpecificHeatCapacity = cp,
|
||
CorrectionTime = tc,
|
||
RSquared = rSquared,
|
||
ValidationMessage = string.Join("\n", validation.Messages),
|
||
CalculationLog = _logBuilder.ToString() // 保存完整日志
|
||
};
|
||
}
|
||
catch (Exception ex)
|
||
{
|
||
LogError($"计算失败: {ex.Message}");
|
||
LogError($"异常堆栈: {ex.StackTrace}");
|
||
return InvalidResult($"计算失败: {ex.Message}");
|
||
}
|
||
}
|
||
|
||
/// <summary>
|
||
/// 计算温度增量ΔT(t) - 国标公式(3)
|
||
/// </summary>
|
||
private double[] CalculateTemperatureIncrements(double[] deltaUArray, double I0)
|
||
{
|
||
LogDebug($"开始计算ΔT(t)，共{deltaUArray.Length}个点");
|
||
double[] deltaT = new double[deltaUArray.Length];
|
||
|
||
for (int i = 0; i < deltaUArray.Length; i++)
|
||
{
|
||
double deltaU = deltaUArray[i];
|
||
|
||
// 国标公式(3): ΔT(t) = (Rs + RL + R0) × ΔU(t) / [(I0 × Rs - ΔU(t)) × α × R0]
|
||
double numerator = (_bridge.Rs + _bridge.RL + _probe.R0) * deltaU;
|
||
double denominator = (I0 * _bridge.Rs - deltaU) * _probe.Alpha * _probe.R0;
|
||
|
||
if (Math.Abs(denominator) < 1e-12)
|
||
{
|
||
LogError($"第{i}点计算ΔT(t)时分母接近0: denominator={denominator:E6}");
|
||
throw new InvalidOperationException($"分母接近0，无法计算ΔT(t)");
|
||
}
|
||
|
||
deltaT[i] = numerator / denominator;
|
||
|
||
// 每100个点记录一次进度
|
||
if (i % 100 == 0 && i > 0)
|
||
{
|
||
LogDebug($"已计算{i}个ΔT点，当前值={deltaT[i]:E6}K");
|
||
}
|
||
}
|
||
|
||
LogDebug($"ΔT(t)计算完成，最后10个值: {string.Join(", ", deltaT.Skip(Math.Max(0, deltaT.Length - 10)).Select(d => d.ToString("E3")))}");
|
||
|
||
return deltaT;
|
||
}
|
||
|
||
/// <summary>
|
||
/// 迭代求解热扩散系数α和校正时间tc
|
||
/// </summary>
|
||
private (double alpha, double tc, string iterationLog) IterateThermalDiffusivity(double[] timeArray, double[] deltaT)
|
||
{
|
||
StringBuilder iterationLog = new StringBuilder();
|
||
iterationLog.AppendLine("迭代过程记录:");
|
||
iterationLog.AppendLine("迭代 | α | tc | 误差 | 改进 | 状态");
|
||
iterationLog.AppendLine("-".PadRight(80, '-'));
|
||
|
||
// 初始猜测值
|
||
double bestAlpha = GetDefaultAlphaBySoilType(); // 常见建筑材料的热扩散系数
|
||
double bestTc = timeArray.Max() * 0.01; // 测试时间的1%作为初始校正时间
|
||
double bestError = double.MaxValue;
|
||
|
||
LogInfo($"迭代初始值: α={bestAlpha:E6} m²/s, tc={bestTc:F4}s");
|
||
|
||
// 使用最小二乘法迭代优化
|
||
int maxIterations = 500;
|
||
double convergenceThreshold = GetDefaultAlphaBySoilType();
|
||
convergenceThreshold = 1e-8;
|
||
int iteration;
|
||
bool converged = false;
|
||
|
||
for (iteration = 0; iteration < maxIterations; iteration++)
|
||
{
|
||
double alphaStep = bestAlpha * 0.1 / (iteration + 1);
|
||
double tcStep = 0.001 / (iteration + 1);
|
||
|
||
var candidates = new List<(double alpha, double tc, double error)>
|
||
{
|
||
(bestAlpha, bestTc, CalculateFitError(timeArray, deltaT, bestAlpha, bestTc)),
|
||
(bestAlpha + alphaStep, bestTc, CalculateFitError(timeArray, deltaT, bestAlpha + alphaStep, bestTc)),
|
||
(bestAlpha - alphaStep, bestTc, CalculateFitError(timeArray, deltaT, bestAlpha - alphaStep, bestTc)),
|
||
(bestAlpha, bestTc + tcStep, CalculateFitError(timeArray, deltaT, bestAlpha, bestTc + tcStep)),
|
||
(bestAlpha, bestTc - tcStep, CalculateFitError(timeArray, deltaT, bestAlpha, bestTc - tcStep))
|
||
};
|
||
|
||
var bestCandidate = candidates.OrderBy(c => c.error).First();
|
||
|
||
double improvement = bestError - bestCandidate.error;
|
||
string status = "搜索";
|
||
|
||
// 收敛条件
|
||
if (Math.Abs(improvement) < convergenceThreshold && iteration > 20)
|
||
{
|
||
status = "收敛";
|
||
converged = true;
|
||
}
|
||
|
||
iterationLog.AppendLine($"{iteration + 1,3} | {bestCandidate.alpha:E6} | {bestCandidate.tc:F6} | {bestCandidate.error:E6} | {improvement:E6} | {status}");
|
||
|
||
if (improvement > 0)
|
||
{
|
||
bestAlpha = bestCandidate.alpha;
|
||
bestTc = bestCandidate.tc;
|
||
bestError = bestCandidate.error;
|
||
}
|
||
|
||
// 每10次迭代记录一次详细状态
|
||
if (iteration % 10 == 0)
|
||
{
|
||
LogDebug($"迭代{iteration}: α={bestAlpha:E6}, tc={bestTc:F6}, 误差={bestError:E6}, 改进={improvement:E6}");
|
||
}
|
||
|
||
// 收敛后停止迭代
|
||
if (converged && iteration > 20)
|
||
{
|
||
LogInfo($"第{iteration + 1}次迭代达到收敛条件，停止迭代");
|
||
break;
|
||
}
|
||
}
|
||
|
||
iterationLog.AppendLine("-".PadRight(80, '-'));
|
||
iterationLog.AppendLine($"总迭代次数: {iteration + 1}次");
|
||
iterationLog.AppendLine($"是否收敛: {(converged ? "是" : "否")}");
|
||
iterationLog.AppendLine($"最终误差: {bestError:E6}");
|
||
iterationLog.AppendLine($"步长调整: α步长={bestAlpha * 0.1 / (iteration + 1):E6}, tc步长={0.001 / (iteration + 1):E6}");
|
||
|
||
if (!converged && iteration >= maxIterations)
|
||
{
|
||
LogWarning($"达到最大迭代次数({maxIterations})仍未收敛");
|
||
iterationLog.AppendLine($"警告: 达到最大迭代次数仍未完全收敛");
|
||
}
|
||
|
||
return (bestAlpha, bestTc, iterationLog.ToString());
|
||
}
|
||
|
||
//private (double alpha, double tc, string iterationLog) IterateThermalDiffusivity(double[] timeArray, double[] deltaT)
|
||
//{
|
||
// StringBuilder iterationLog = new StringBuilder();
|
||
// iterationLog.AppendLine("迭代过程记录:");
|
||
// iterationLog.AppendLine("迭代 | α | tc | 误差 | 改进 | 状态");
|
||
// iterationLog.AppendLine("-".PadRight(80, '-'));
|
||
|
||
// // 强制从更大的α值开始搜索
|
||
// double r = _probe.RadiusMM / 1000.0;
|
||
// double tmax = timeArray.Max();
|
||
|
||
// // 基于τ²=0.8估算初始α（之前可能估算偏小）
|
||
// double initialAlpha = 0.8 * r * r / (tmax * 0.99); // τ²=0.8，假设tc≈0
|
||
|
||
// // 关键修改：强制初始α不小于1e-6
|
||
// initialAlpha = Math.Max(initialAlpha, 2.0e-6);
|
||
// initialAlpha = Math.Min(initialAlpha, 3.0e-6);
|
||
|
||
// double bestAlpha = initialAlpha;
|
||
// double bestTc = tmax * 0.01;
|
||
// double bestError = CalculateFitError(timeArray, deltaT, bestAlpha, bestTc);
|
||
|
||
// LogInfo($"强制初始α: α={bestAlpha:E6} m²/s (基于τ²=0.8估算), tc={bestTc:F4}s");
|
||
|
||
// int maxIterations = 200;
|
||
// double convergenceThreshold = 1e-8;
|
||
// int iteration;
|
||
// bool converged = false;
|
||
|
||
// // 记录最佳解的改进历史
|
||
// List<double> errorHistory = new List<double> { bestError };
|
||
|
||
// for (iteration = 0; iteration < maxIterations; iteration++)
|
||
// {
|
||
// // 动态步长：初期探索大，后期精细调整
|
||
// double alphaStep = bestAlpha * (0.3 / (1 + iteration * 0.05)); // 衰减更慢
|
||
// double tcStep = 0.01 / (1 + iteration * 0.05);
|
||
|
||
// // 确保最小步长
|
||
// alphaStep = Math.Max(alphaStep, bestAlpha * 0.05);
|
||
// tcStep = Math.Max(tcStep, 0.001);
|
||
|
||
// // 生成候选解
|
||
// var candidates = new List<(double alpha, double tc, double error)>
|
||
//{
|
||
// (bestAlpha, bestTc, bestError),
|
||
// (bestAlpha + alphaStep, bestTc, CalculateFitError(timeArray, deltaT, bestAlpha + alphaStep, bestTc)),
|
||
// (bestAlpha - alphaStep, bestTc, CalculateFitError(timeArray, deltaT, bestAlpha - alphaStep, bestTc)),
|
||
// (bestAlpha, bestTc + tcStep, CalculateFitError(timeArray, deltaT, bestAlpha, bestTc + tcStep)),
|
||
// (bestAlpha, bestTc - tcStep, CalculateFitError(timeArray, deltaT, bestAlpha, bestTc - tcStep)),
|
||
// // 强制探索更大的α值
|
||
// (bestAlpha * 1.5, bestTc, CalculateFitError(timeArray, deltaT, bestAlpha * 1.5, bestTc)),
|
||
// (bestAlpha * 0.67, bestTc, CalculateFitError(timeArray, deltaT, bestAlpha * 0.67, bestTc)),
|
||
//};
|
||
|
||
// var bestCandidate = candidates.OrderBy(c => c.error).First();
|
||
// double improvement = bestError - bestCandidate.error;
|
||
|
||
// string status = "搜索";
|
||
|
||
// // 收敛条件：连续5次改进小于阈值
|
||
// errorHistory.Add(bestCandidate.error);
|
||
// if (errorHistory.Count > 5)
|
||
// {
|
||
// double recentImprovement = errorHistory[errorHistory.Count - 6] - errorHistory[errorHistory.Count - 1];
|
||
// if (Math.Abs(recentImprovement) < convergenceThreshold * 5 && iteration > 20)
|
||
// {
|
||
// status = "收敛";
|
||
// converged = true;
|
||
// }
|
||
// }
|
||
|
||
// iterationLog.AppendLine($"{iteration + 1,3} | {bestCandidate.alpha:E6} | {bestCandidate.tc:F6} | {bestCandidate.error:E6} | {improvement:E6} | {status}");
|
||
|
||
// if (improvement > 0 || (improvement == 0 && bestCandidate.error < bestError))
|
||
// {
|
||
// bestAlpha = bestCandidate.alpha;
|
||
// bestTc = bestCandidate.tc;
|
||
// bestError = bestCandidate.error;
|
||
// }
|
||
|
||
// // 每10次迭代记录一次
|
||
// if (iteration % 10 == 0)
|
||
// {
|
||
// LogDebug($"迭代{iteration}: α={bestAlpha:E6}, tc={bestTc:F6}, 误差={bestError:E6}");
|
||
|
||
// // 如果连续多次没有改进，尝试跳跃
|
||
// if (iteration > 30 && errorHistory.Count > 10)
|
||
// {
|
||
// double lastImprovement = errorHistory[errorHistory.Count - 11] - errorHistory[errorHistory.Count - 1];
|
||
// if (Math.Abs(lastImprovement) < 1e-6)
|
||
// {
|
||
// // 随机扰动
|
||
// double randomFactor = 0.8 + (new Random().NextDouble() * 0.4);
|
||
// bestAlpha *= randomFactor;
|
||
// LogDebug($"迭代{iteration}: 停滞检测，随机扰动α乘以{randomFactor:F3}");
|
||
// }
|
||
// }
|
||
// }
|
||
|
||
// if (converged && iteration > 20)
|
||
// {
|
||
// LogInfo($"第{iteration + 1}次迭代达到收敛条件");
|
||
// break;
|
||
// }
|
||
// }
|
||
|
||
// iterationLog.AppendLine("-".PadRight(80, '-'));
|
||
// iterationLog.AppendLine($"总迭代次数: {iteration + 1}次");
|
||
// iterationLog.AppendLine($"是否收敛: {(converged ? "是" : "否")}");
|
||
// iterationLog.AppendLine($"最终误差: {bestError:E6}");
|
||
|
||
// // 验证α是否合理
|
||
// if (bestAlpha < 5e-7)
|
||
// {
|
||
// LogWarning($"警告：热扩散系数α={bestAlpha:E6}可能过小！建议：");
|
||
// LogWarning($" 1. 检查样品是否真的导热很差");
|
||
// LogWarning($" 2. 检查探头参数R0、α是否正确");
|
||
// LogWarning($" 3. 尝试增加测试时间以获得更大的τ²");
|
||
// }
|
||
|
||
// return (bestAlpha, bestTc, iterationLog.ToString());
|
||
//}
|
||
|
||
/// <summary>
|
||
/// 计算拟合误差 - 使用过原点回归的残差平方和
|
||
/// </summary>
|
||
private double CalculateFitError(double[] timeArray, double[] deltaT, double alpha, double tc)
|
||
{
|
||
double r = _probe.RadiusMM / 1000.0;
|
||
var dTauList = new List<double>();
|
||
var deltaTList = new List<double>();
|
||
|
||
// 收集有效数据点
|
||
for (int i = 0; i < timeArray.Length; i++)
|
||
{
|
||
if (timeArray[i] > tc)
|
||
{
|
||
double tau = CalculateTau(timeArray[i], tc, alpha, r);
|
||
|
||
// 检查τ是否在合理范围内 - 修复：使用正确的τ范围 [√0.3, 1] = [0.5477, 1]
|
||
if (tau >= Math.Sqrt(TAU_MAX_SQUARED_LOWER) && tau <= Math.Sqrt(TAU_MAX_SQUARED_UPPER))
|
||
{
|
||
// 边界检查
|
||
if (tau < 0.01 || tau > 3.0) continue;
|
||
|
||
try
|
||
{
|
||
double dTau = _tauDInterpolation.Interpolate(tau);
|
||
dTauList.Add(dTau);
|
||
deltaTList.Add(deltaT[i]);
|
||
}
|
||
catch (Exception ex)
|
||
{
|
||
LogDebug($"第{i}点插值失败: τ={tau:F4}, 错误={ex.Message}");
|
||
continue;
|
||
}
|
||
}
|
||
}
|
||
}
|
||
|
||
if (dTauList.Count < 10)
|
||
{
|
||
LogDebug($"误差计算: 有效数据点不足({dTauList.Count})，返回最大误差");
|
||
return double.MaxValue;
|
||
}
|
||
|
||
// 过原点线性回归: ΔT = k * D(τ)
|
||
double numerator = 0;
|
||
double denominator = 0;
|
||
|
||
for (int i = 0; i < dTauList.Count; i++)
|
||
{
|
||
numerator += dTauList[i] * deltaTList[i];
|
||
denominator += dTauList[i] * dTauList[i];
|
||
}
|
||
|
||
if (Math.Abs(denominator) < 1e-12)
|
||
{
|
||
LogDebug("误差计算: 分母接近0，返回最大误差");
|
||
return double.MaxValue;
|
||
}
|
||
|
||
double slope = numerator / denominator; // k = Σ(D*ΔT) / Σ(D²)
|
||
|
||
// 计算均方根误差 (RMSE)
|
||
double totalError = 0;
|
||
for (int i = 0; i < dTauList.Count; i++)
|
||
{
|
||
double predicted = slope * dTauList[i]; // 过原点预测值
|
||
double error = deltaTList[i] - predicted;
|
||
totalError += error * error;
|
||
}
|
||
|
||
double rmse = Math.Sqrt(totalError / dTauList.Count);
|
||
double relativeRmse = rmse / Math.Abs(deltaTList.Average()); // 相对误差
|
||
|
||
LogDebug($"误差计算: 使用{dTauList.Count}个点, 斜率={slope:E6}, RMSE={rmse:E6}, 相对误差={relativeRmse:P2}");
|
||
|
||
return relativeRmse; // 返回相对误差
|
||
}
|
||
///// <summary>
|
||
///// 计算拟合误差 - 增加对α变化的敏感性
|
||
///// </summary>
|
||
//private double CalculateFitError(double[] timeArray, double[] deltaT, double alpha, double tc)
|
||
//{
|
||
// double r = _probe.RadiusMM / 1000.0;
|
||
// var dTauList = new List<double>();
|
||
// var deltaTList = new List<double>();
|
||
|
||
// // 收集有效数据点
|
||
// for (int i = 0; i < timeArray.Length; i++)
|
||
// {
|
||
// if (timeArray[i] > tc)
|
||
// {
|
||
// double tau = CalculateTau(timeArray[i], tc, alpha, r);
|
||
|
||
// // 检查τ是否在国标范围内 [√0.3, 1] = [0.5477, 1]
|
||
// if (tau >= Math.Sqrt(TAU_MAX_SQUARED_LOWER) && tau <= Math.Sqrt(TAU_MAX_SQUARED_UPPER))
|
||
// {
|
||
// if (tau < 0.01 || tau > 3.0) continue;
|
||
|
||
// try
|
||
// {
|
||
// double dTau = _tauDInterpolation.Interpolate(tau);
|
||
// dTauList.Add(dTau);
|
||
// deltaTList.Add(deltaT[i]);
|
||
// }
|
||
// catch { continue; }
|
||
// }
|
||
// }
|
||
// }
|
||
|
||
// if (dTauList.Count < 10)
|
||
// {
|
||
// return double.MaxValue; // 返回最大误差
|
||
// }
|
||
|
||
// // 过原点线性回归
|
||
// double numerator = 0, denominator = 0;
|
||
// for (int i = 0; i < dTauList.Count; i++)
|
||
// {
|
||
// numerator += dTauList[i] * deltaTList[i];
|
||
// denominator += dTauList[i] * dTauList[i];
|
||
// }
|
||
|
||
// if (Math.Abs(denominator) < 1e-12) return double.MaxValue;
|
||
|
||
// double slope = numerator / denominator;
|
||
|
||
// // 关键修改：增加惩罚项，鼓励α在合理范围内
|
||
// double rmse = 0;
|
||
// for (int i = 0; i < dTauList.Count; i++)
|
||
// {
|
||
// double predicted = slope * dTauList[i];
|
||
// double error = deltaTList[i] - predicted;
|
||
// rmse += error * error;
|
||
// }
|
||
// rmse = Math.Sqrt(rmse / dTauList.Count);
|
||
|
||
// // 修改1：对α过小的情况增加惩罚
|
||
// double alphaPenalty = 0;
|
||
// if (alpha < 5e-7) // 如果α小于0.5e-6，增加惩罚
|
||
// {
|
||
// alphaPenalty = (5e-7 - alpha) * 100; // 惩罚系数
|
||
// }
|
||
|
||
// // 修改2：对斜率合理性增加检查（根据物理意义）
|
||
// double expectedSlopeMin = _test.P0 / (PI_POW_1_5 * r * 3.0); // λ=3时的最小斜率
|
||
// double expectedSlopeMax = _test.P0 / (PI_POW_1_5 * r * 0.1); // λ=0.1时的最大斜率
|
||
|
||
// double slopePenalty = 0;
|
||
// if (slope < expectedSlopeMin || slope > expectedSlopeMax)
|
||
// {
|
||
// slopePenalty = 1.0; // 显著惩罚
|
||
// }
|
||
|
||
// // 综合误差 = RMSE + 惩罚项
|
||
// double totalError = rmse + alphaPenalty + slopePenalty;
|
||
|
||
// LogDebug($"误差计算: α={alpha:E6}, 使用{dTauList.Count}点, RMSE={rmse:E3}, α惩罚={alphaPenalty:E3}, 斜率惩罚={slopePenalty:E3}, 总误差={totalError:E3}");
|
||
|
||
// return totalError;
|
||
//}
|
||
|
||
|
||
/// <summary>
|
||
/// 计算导热系数λ - 国标公式(4)，使用国标范围内的数据
|
||
/// </summary>
|
||
private double CalculateThermalConductivity(double[] timeArray, double[] deltaT, double alpha, double tc)
|
||
{
|
||
LogInfo("开始计算导热系数λ");
|
||
double r = _probe.RadiusMM / 1000.0;
|
||
var validPoints = new List<(double dTau, double deltaT)>();
|
||
|
||
// 收集有效数据点（严格按国标τ范围筛选）
|
||
int validCount = 0;
|
||
int totalCount = 0;
|
||
int tGreaterTcCount = 0;
|
||
|
||
LogDebug("筛选有效数据点用于线性回归:");
|
||
for (int i = 0; i < timeArray.Length; i++)
|
||
{
|
||
totalCount++;
|
||
|
||
// 只使用校正后的数据（t > t_c）
|
||
if (timeArray[i] > tc)
|
||
{
|
||
tGreaterTcCount++;
|
||
double tau = CalculateTau(timeArray[i], tc, alpha, r);
|
||
|
||
// 只使用τ在国标要求范围内的数据点
|
||
if (tau >= Math.Sqrt(TAU_MAX_SQUARED_LOWER) && tau <= Math.Sqrt(TAU_MAX_SQUARED_UPPER))
|
||
{
|
||
try
|
||
{
|
||
double dTau = _tauDInterpolation.Interpolate(tau);
|
||
validPoints.Add((dTau, deltaT[i]));
|
||
validCount++;
|
||
|
||
// 每50个有效点记录一次
|
||
if (validCount % 50 == 0)
|
||
{
|
||
LogDebug($"已找到{validCount}个有效点，当前点: t={timeArray[i]:F4}s, τ={tau:F4}, D(τ)={dTau:F6}, ΔT={deltaT[i]:E6}K");
|
||
}
|
||
}
|
||
catch (Exception ex)
|
||
{
|
||
LogDebug($"第{i}点插值失败: τ={tau:F4}, 错误={ex.Message}");
|
||
continue;
|
||
}
|
||
}
|
||
}
|
||
}
|
||
|
||
LogInfo($"数据点统计: 总数={totalCount}, t>tc点数={tGreaterTcCount}, τ有效点数={validCount}");
|
||
|
||
if (validPoints.Count < 10)
|
||
{
|
||
LogError($"有效数据点不足({validPoints.Count})，无法计算导热系数");
|
||
throw new InvalidOperationException(
|
||
$"有效数据点不足({validPoints.Count})，无法计算导热系数\n" +
|
||
$"国标要求：在τ范围[{TAU_MAX_SQUARED_LOWER}, {TAU_MAX_SQUARED_UPPER}]内应有足够数据点");
|
||
}
|
||
|
||
LogInfo($"使用{validPoints.Count}个有效数据点进行线性回归");
|
||
|
||
// 线性回归：ΔT(τ) = [P₀ / (π^(3/2) × r × λ)] × D(τ)
|
||
double[] dTauArray = validPoints.Select(p => p.dTau).ToArray();
|
||
double[] deltaTArray = validPoints.Select(p => p.deltaT).ToArray();
|
||
|
||
LogDebug($"D(τ)范围: [{dTauArray.Min():F6}, {dTauArray.Max():F6}]");
|
||
LogDebug($"ΔT范围: [{deltaTArray.Min():E6}, {deltaTArray.Max():E6}]");
|
||
|
||
var (slope, intercept) = LinearRegression(dTauArray, deltaTArray, true);
|
||
LogInfo($"线性回归结果: 斜率={slope:E6}, 截距={intercept:E6}, R={CalculateCorrelation(dTauArray, deltaTArray):F6}");
|
||
|
||
// 检查斜率是否合理
|
||
if (Math.Abs(slope) < 1e-12)
|
||
{
|
||
LogError($"回归斜率接近0 ({slope:E6})，计算结果不可靠");
|
||
throw new InvalidOperationException("回归斜率接近0，计算结果不可靠");
|
||
}
|
||
|
||
// 计算导热系数 λ = P₀ / (π^(3/2) × r × slope) - 国标公式(4)
|
||
double lambda = _test.P0 / (PI_POW_1_5 * r * slope);
|
||
LogInfo($"导热系数计算: P0={_test.P0}W, r={r}m, π^(3/2)={PI_POW_1_5:F6}, slope={slope:E6}");
|
||
LogInfo($"最终导热系数: λ={lambda:F6} W/(m·K)");
|
||
|
||
return lambda;
|
||
}
|
||
|
||
/// <summary>
|
||
/// 计算无量纲时间τ - 国标公式(5)
|
||
/// τ = √[(t - t_c) / (r²/a)]
|
||
/// 等价于：τ = √[(t - t_c) * a / r²]
|
||
/// </summary>
|
||
private double CalculateTau(double time, double tc, double alpha, double r)
|
||
{
|
||
if (time <= tc)
|
||
{
|
||
return 0;
|
||
}
|
||
|
||
// 国标公式(5): τ = √[(t - t_c) / (r²/a)]
|
||
double theta = (r * r) / alpha; // 特征时间 θ = r²/a
|
||
double tau = Math.Sqrt((time - tc) / theta);
|
||
|
||
return tau;
|
||
}
|
||
|
||
/// <summary>
|
||
/// 计算比热容Cp - Cp = λ / (ρ × α)
|
||
/// </summary>
|
||
private double CalculateSpecificHeatCapacity(double lambda, double alpha)
|
||
{
|
||
LogInfo($"开始计算比热容: λ={lambda:F6}, α={alpha:E6}, ρ={_test.SampleDensity}");
|
||
|
||
if (alpha <= 0 || _test.SampleDensity <= 0)
|
||
{
|
||
LogError($"参数无效: α={alpha:E6}, ρ={_test.SampleDensity}");
|
||
throw new InvalidOperationException("热扩散系数或密度无效，无法计算比热容");
|
||
}
|
||
|
||
double cp = lambda / (_test.SampleDensity * alpha);
|
||
LogInfo($"比热容计算: Cp = {lambda:F6} / ({_test.SampleDensity} × {alpha:E6}) = {cp:F2} J/(kg·K)");
|
||
|
||
return cp;
|
||
}
|
||
|
||
/// <summary>
|
||
/// 验证测试结果 - 严格遵循国标5.3.4节要求
|
||
/// </summary>
|
||
private ValidationResult ValidateTestResults(double[] timeArray, double alpha, double tc, double tauMaxSquared)
|
||
{
|
||
LogInfo("开始测试结果验证");
|
||
var result = new ValidationResult { IsValid = true };
|
||
double r = _probe.RadiusMM / 1000.0; // 转换为米
|
||
double tmax = timeArray.Max();
|
||
|
||
// 1. 验证τ_max²是否满足国标要求
|
||
if (tauMaxSquared < TAU_MAX_SQUARED_LOWER)
|
||
{
|
||
result.IsValid = false;
|
||
string msg = $"τ_max²({tauMaxSquared:F3}) < {TAU_MAX_SQUARED_LOWER}，测试时间不足，结果无效";
|
||
LogError(msg);
|
||
result.Messages.Add(msg);
|
||
}
|
||
else if (tauMaxSquared > TAU_MAX_SQUARED_UPPER)
|
||
{
|
||
result.IsValid = false;
|
||
string msg = $"τ_max²({tauMaxSquared:F3}) > {TAU_MAX_SQUARED_UPPER}，测试时间过长，结果无效";
|
||
LogError(msg);
|
||
result.Messages.Add(msg);
|
||
}
|
||
else
|
||
{
|
||
string msg = $"τ_max²({tauMaxSquared:F3})满足国标要求[{TAU_MAX_SQUARED_LOWER}, {TAU_MAX_SQUARED_UPPER}]";
|
||
LogInfo(msg);
|
||
result.Messages.Add(msg);
|
||
}
|
||
|
||
// 2. 计算探测深度 ΔP_probe = 2√(a·t_max) （国标5.3.4）
|
||
double probingDepth = 2 * Math.Sqrt(alpha * tmax);
|
||
double probingDepthRatio = probingDepth / r;
|
||
LogInfo($"探测深度计算: 2×√({alpha:E6}×{tmax:F2}) = {probingDepth * 1000:F2}mm, 探头半径={r * 1000:F2}mm, 比值={probingDepthRatio:F2}");
|
||
|
||
// 3. 验证探测深度是否满足国标要求
|
||
if (probingDepthRatio < PROBING_DEPTH_RATIO_LOWER)
|
||
{
|
||
result.IsValid = false;
|
||
string msg = $"探测深度/探头半径({probingDepthRatio:F2}) < {PROBING_DEPTH_RATIO_LOWER}，样品厚度可能不足";
|
||
LogError(msg);
|
||
result.Messages.Add(msg);
|
||
}
|
||
else if (probingDepthRatio > PROBING_DEPTH_RATIO_UPPER)
|
||
{
|
||
result.IsValid = false;
|
||
string msg = $"探测深度/探头半径({probingDepthRatio:F2}) > {PROBING_DEPTH_RATIO_UPPER}，可能超出测试范围";
|
||
LogError(msg);
|
||
result.Messages.Add(msg);
|
||
}
|
||
else
|
||
{
|
||
string msg = $"探测深度/探头半径({probingDepthRatio:F2})满足国标要求[{PROBING_DEPTH_RATIO_LOWER}, {PROBING_DEPTH_RATIO_UPPER}]";
|
||
LogInfo(msg);
|
||
result.Messages.Add(msg);
|
||
}
|
||
|
||
// 4. 验证数据点数量（国标5.3.3.3要求采集次数大于100次）
|
||
if (timeArray.Length < MIN_DATA_COUNT)
|
||
{
|
||
string msg = $"警告：采集次数({timeArray.Length})少于国标要求的{MIN_DATA_COUNT}次";
|
||
LogWarning(msg);
|
||
result.Messages.Add(msg);
|
||
}
|
||
else
|
||
{
|
||
string msg = $"采集次数({timeArray.Length})满足国标要求(≥{MIN_DATA_COUNT}次)";
|
||
LogInfo(msg);
|
||
result.Messages.Add(msg);
|
||
}
|
||
|
||
// 5. 验证数据采集间隔（国标5.3.3.3要求不小于0.1s）
|
||
if (timeArray.Length > 1)
|
||
{
|
||
double minInterval = timeArray[1] - timeArray[0];
|
||
for (int i = 2; i < timeArray.Length; i++)
|
||
{
|
||
double interval = timeArray[i] - timeArray[i - 1];
|
||
if (interval < minInterval) minInterval = interval;
|
||
}
|
||
|
||
if (minInterval < MIN_TIME_INTERVAL)
|
||
{
|
||
string msg = $"警告：数据采集间隔({minInterval:F3}s)小于国标要求的{MIN_TIME_INTERVAL}s";
|
||
LogWarning(msg);
|
||
result.Messages.Add(msg);
|
||
}
|
||
else
|
||
{
|
||
string msg = $"数据采集间隔({minInterval:F3}s)满足国标要求(≥{MIN_TIME_INTERVAL}s)";
|
||
LogInfo(msg);
|
||
result.Messages.Add(msg);
|
||
}
|
||
}
|
||
|
||
// 6. 生成详细验证报告
|
||
result.Messages.Insert(0, $"【国标GB/T 32064-2015测试结果有效性验证】");
|
||
result.Messages.Insert(1, $"测试总时间: {tmax:F2}s，校正时间: {tc:F4}s");
|
||
result.Messages.Insert(2, $"热扩散系数α: {alpha:E6} m²/s");
|
||
result.Messages.Insert(3, $"探头半径r: {r * 1000:F2}mm，探测深度: {probingDepth * 1000:F2}mm");
|
||
result.Messages.Insert(4, $"验证依据：5.3.4节 测试结果有效性验证");
|
||
|
||
LogInfo($"测试结果验证完成: {(result.IsValid ? "有效" : "无效")}");
|
||
|
||
return result;
|
||
}
|
||
|
||
/// <summary>
|
||
/// 计算τ_max²
|
||
/// </summary>
|
||
private double CalculateTauMaxSquared(double tmax, double tc, double alpha, double r)
|
||
{
|
||
LogDebug($"计算τ_max²: tmax={tmax:F4}, tc={tc:F4}, α={alpha:E6}, r={r:F6}");
|
||
|
||
// 国标公式：τ_max² = (t_max - t_c) × a / r²
|
||
if (r <= 0) throw new ArgumentException("探头半径必须大于0");
|
||
if (tmax <= tc)
|
||
{
|
||
LogWarning($"tmax({tmax:F4}) <= tc({tc:F4})，返回0");
|
||
return 0;
|
||
}
|
||
|
||
//double tauMaxSquared = (tmax - tc) * alpha / (r * r);
|
||
double tauMax = Math.Sqrt((tmax - tc) * alpha / (r * r));
|
||
double tauMaxSquared = tauMax * tauMax; // τ² = τ×τ
|
||
LogDebug($"τ_max²计算结果: ({tmax}-{tc})×{alpha:E6}/{r * r:E6} = {tauMaxSquared:F4}");
|
||
|
||
return tauMaxSquared;
|
||
}
|
||
|
||
/// <summary>
|
||
/// 计算R²拟合优度
|
||
/// </summary>
|
||
private double CalculateRSquared(double[] timeArray, double[] deltaT, double alpha, double tc, double lambda)
|
||
{
|
||
LogInfo("开始计算R²拟合优度");
|
||
var observed = new List<double>();
|
||
var predicted = new List<double>();
|
||
double r = _probe.RadiusMM / 1000.0;
|
||
int validCount = 0;
|
||
|
||
for (int i = 0; i < timeArray.Length; i++)
|
||
{
|
||
if (timeArray[i] > tc)
|
||
{
|
||
double tau = CalculateTau(timeArray[i], tc, alpha, r);
|
||
if (tau >= TAU_MAX_SQUARED_LOWER && tau <= TAU_MAX_SQUARED_UPPER)
|
||
{
|
||
try
|
||
{
|
||
double dTau = _tauDInterpolation.Interpolate(tau);
|
||
double theoretical = _test.P0 * dTau / (PI_POW_1_5 * r * lambda);
|
||
|
||
observed.Add(deltaT[i]);
|
||
predicted.Add(theoretical);
|
||
validCount++;
|
||
}
|
||
catch
|
||
{
|
||
continue;
|
||
}
|
||
}
|
||
}
|
||
}
|
||
|
||
LogInfo($"R²计算使用{validCount}个有效数据点");
|
||
|
||
if (observed.Count < 10)
|
||
{
|
||
LogWarning($"R²计算数据点不足({observed.Count})，返回0");
|
||
return 0;
|
||
}
|
||
|
||
double meanObserved = observed.Average();
|
||
double ssTotal = observed.Sum(o => Math.Pow(o - meanObserved, 2));
|
||
double ssResidual = observed.Zip(predicted, (o, p) => Math.Pow(o - p, 2)).Sum();
|
||
|
||
if (ssResidual > ssTotal)
|
||
{
|
||
LogWarning($"SSR({ssResidual:E6}) > SST({ssTotal:E6})，强制设为0");
|
||
return 0;
|
||
}
|
||
LogDebug($"R²计算: 平均值={meanObserved:E6}, SST={ssTotal:E6}, SSR={ssResidual:E6}");
|
||
|
||
if (Math.Abs(ssTotal) < 1e-12)
|
||
{
|
||
LogWarning("SST接近0，返回0");
|
||
return 0;
|
||
}
|
||
|
||
double rSquared = 1 - (ssResidual / ssTotal);
|
||
LogInfo($"R²计算结果: {rSquared:F6} (SST={ssTotal:E6}, SSR={ssResidual:E6})");
|
||
|
||
return rSquared;
|
||
}
|
||
#endregion
|
||
|
||
#region 辅助方法
|
||
private CalculationResult InvalidResult(string message)
|
||
{
|
||
LogError($"返回无效结果: {message}");
|
||
return new CalculationResult
|
||
{
|
||
IsValid = false,
|
||
ValidationMessage = message,
|
||
CalculationLog = _logBuilder.ToString(),
|
||
ThermalConductivity = 0,
|
||
ThermalDiffusivity = 0,
|
||
SpecificHeatCapacity = 0,
|
||
CorrectionTime = 0,
|
||
RSquared = 0
|
||
};
|
||
}
|
||
|
||
private bool ValidateInputData(double[] timeArray, double[] deltaUArray)
|
||
{
|
||
LogDebug("开始验证输入数据");
|
||
|
||
if (timeArray == null || deltaUArray == null)
|
||
{
|
||
LogError("时间和电压数据不能为空");
|
||
throw new ArgumentNullException("时间和电压数据不能为空");
|
||
}
|
||
|
||
if (timeArray.Length != deltaUArray.Length)
|
||
{
|
||
LogError($"数据长度不一致: 时间{timeArray.Length}点, 电压{deltaUArray.Length}点");
|
||
throw new ArgumentException("时间和电压数据长度不一致");
|
||
}
|
||
|
||
if (timeArray.Length < 20)
|
||
{
|
||
LogError($"数据点数量({timeArray.Length})不足，至少需要20个点");
|
||
throw new ArgumentException($"数据点数量({timeArray.Length})不足，至少需要20个点");
|
||
}
|
||
|
||
// 检查时间单调递增
|
||
for (int i = 1; i < timeArray.Length; i++)
|
||
{
|
||
if (timeArray[i] <= timeArray[i - 1])
|
||
{
|
||
LogError($"时间数据非单调递增: 第{i - 1}点={timeArray[i - 1]}, 第{i}点={timeArray[i]}");
|
||
throw new ArgumentException($"时间数据必须严格单调递增，第{i}点不符合要求");
|
||
}
|
||
}
|
||
|
||
// 统计信息
|
||
double timeSpan = timeArray[timeArray.Length - 1] - timeArray[0];
|
||
double avgInterval = timeSpan / (timeArray.Length - 1);
|
||
LogInfo($"数据验证通过: 时间范围[{timeArray[0]:F4}s, {timeArray[timeArray.Length - 1]:F4}s], 时长={timeSpan:F2}s");
|
||
LogInfo($"数据统计: 点数={timeArray.Length}, 平均间隔={avgInterval:F4}s");
|
||
LogInfo($"电压范围: [{deltaUArray.Min():E6}V, {deltaUArray.Max():E6}V], 平均值={deltaUArray.Average():E6}V");
|
||
|
||
return true;
|
||
}
|
||
|
||
private (double slope, double intercept) LinearRegression(double[] x, double[] y, bool forceOrigin = true)
|
||
{
|
||
LogDebug($"线性回归: 输入{x.Length}个点");
|
||
if (x.Length != y.Length || x.Length < 2)
|
||
{
|
||
LogError($"线性回归参数错误: x长度={x.Length}, y长度={y.Length}");
|
||
throw new ArgumentException("线性回归需要至少2个数据点且x、y长度相等");
|
||
}
|
||
|
||
if (forceOrigin)
|
||
{
|
||
// 过原点回归：y = kx
|
||
double numerator = 0;
|
||
double denominator = 0;
|
||
|
||
for (int i = 0; i < x.Length; i++)
|
||
{
|
||
numerator += x[i] * y[i];
|
||
denominator += x[i] * x[i];
|
||
}
|
||
|
||
if (Math.Abs(denominator) < 1e-12)
|
||
{
|
||
LogError("数据方差太小，无法进行线性回归");
|
||
throw new InvalidOperationException("数据方差太小，无法进行线性回归");
|
||
}
|
||
|
||
double slope = numerator / denominator;
|
||
double intercept = 0;
|
||
|
||
LogDebug($"过原点回归: 斜率={slope:E6}");
|
||
return (slope, intercept);
|
||
}
|
||
else
|
||
{
|
||
double xAvg = x.Average();
|
||
double yAvg = y.Average();
|
||
LogDebug($"平均值: x̄={xAvg:E6}, ȳ={yAvg:E6}");
|
||
|
||
double numerator = 0;
|
||
double denominator = 0;
|
||
|
||
for (int i = 0; i < x.Length; i++)
|
||
{
|
||
double xDiff = x[i] - xAvg;
|
||
double yDiff = y[i] - yAvg;
|
||
numerator += xDiff * yDiff;
|
||
denominator += xDiff * xDiff;
|
||
}
|
||
|
||
LogDebug($"回归计算: 分子={numerator:E6}, 分母={denominator:E6}");
|
||
|
||
if (Math.Abs(denominator) < 1e-12)
|
||
{
|
||
LogError("数据方差太小，无法进行线性回归");
|
||
throw new InvalidOperationException("数据方差太小，无法进行线性回归");
|
||
}
|
||
|
||
double slope = numerator / denominator;
|
||
double intercept = yAvg - slope * xAvg;
|
||
|
||
LogDebug($"回归结果: 斜率={slope:E6}, 截距={intercept:E6}");
|
||
|
||
return (slope, intercept);
|
||
}
|
||
}
|
||
|
||
/// <summary>
|
||
/// 计算相关系数
|
||
/// </summary>
|
||
private double CalculateCorrelation(double[] x, double[] y)
|
||
{
|
||
if (x.Length != y.Length || x.Length < 2) return 0;
|
||
|
||
double xAvg = x.Average();
|
||
double yAvg = y.Average();
|
||
|
||
double numerator = 0;
|
||
double xSumSq = 0;
|
||
double ySumSq = 0;
|
||
|
||
for (int i = 0; i < x.Length; i++)
|
||
{
|
||
double xDiff = x[i] - xAvg;
|
||
double yDiff = y[i] - yAvg;
|
||
numerator += xDiff * yDiff;
|
||
xSumSq += xDiff * xDiff;
|
||
ySumSq += yDiff * yDiff;
|
||
}
|
||
|
||
if (xSumSq == 0 || ySumSq == 0) return 0;
|
||
|
||
return numerator / Math.Sqrt(xSumSq * ySumSq);
|
||
}
|
||
#endregion
|
||
|
||
#region 日志记录方法
|
||
private void LogInfo(string message)
|
||
{
|
||
string logEntry = $"[INFO] {DateTime.Now:HH:mm:ss.fff} - {message}";
|
||
_logBuilder.AppendLine(logEntry);
|
||
// 这里也可以输出到控制台或文件
|
||
// Console.WriteLine(logEntry);
|
||
}
|
||
|
||
private void LogDebug(string message)
|
||
{
|
||
string logEntry = $"[DEBUG] {DateTime.Now:HH:mm:ss.fff} - {message}";
|
||
_logBuilder.AppendLine(logEntry);
|
||
// 调试信息可根据需要启用或禁用
|
||
// Console.WriteLine(logEntry);
|
||
}
|
||
|
||
private void LogWarning(string message)
|
||
{
|
||
string logEntry = $"[WARNING] {DateTime.Now:HH:mm:ss.fff} - {message}";
|
||
_logBuilder.AppendLine(logEntry);
|
||
// Console.WriteLine(logEntry);
|
||
}
|
||
|
||
private void LogError(string message)
|
||
{
|
||
string logEntry = $"[ERROR] {DateTime.Now:HH:mm:ss.fff} - {message}";
|
||
_logBuilder.AppendLine(logEntry);
|
||
// Console.WriteLine(logEntry);
|
||
}
|
||
#endregion
|
||
|
||
#region τ-D(τ)插值表初始化
|
||
|
||
private void InitializeTauDTable()
|
||
{
|
||
LogInfo("初始化τ-D(τ)插值表");
|
||
|
||
// 根据国标GB/T 32064-2015，瞬态平面热源法的D(τ)函数
|
||
// D(τ) = [√(τ²+1) - 1] / τ 或 D(τ) = [√(τ²+1) - τ] / [√(τ²+1) + τ]
|
||
// 需要确认哪个是正确的公式！
|
||
|
||
var tauList = new List<double>();
|
||
var dList = new List<double>();
|
||
|
||
LogDebug("生成τ-D(τ)数据点");
|
||
|
||
// 先计算并记录两种公式的差异
|
||
LogDebug("τ-D(τ)公式验证:");
|
||
for (double tau = 0.5; tau <= 2.5; tau += 0.5)
|
||
{
|
||
double d1 = (Math.Sqrt(tau * tau + 1.0) - 1.0) / tau;
|
||
double d2 = (Math.Sqrt(tau * tau + 1.0) - tau) / (Math.Sqrt(tau * tau + 1.0) + tau);
|
||
LogDebug($"τ={tau:F2}: 公式1(D1)={d1:F6}, 公式2(D2)={d2:F6}, 比值={d1 / d2:F3}");
|
||
}
|
||
|
||
for (double tau = 0.01; tau <= 3.0; tau += 0.01)
|
||
{
|
||
tauList.Add(tau);
|
||
double d = (Math.Sqrt(tau * tau + 1.0) - 1.0) / tau;
|
||
dList.Add(d);
|
||
}
|
||
|
||
double[] tauValues = tauList.ToArray();
|
||
double[] dValues = dList.ToArray();
|
||
|
||
_tauDInterpolation = CubicSpline.InterpolateAkimaSorted(tauValues, dValues);
|
||
LogInfo($"τ-D(τ)插值表初始化完成，共{tauValues.Length}个点，τ范围[{tauValues[0]:F2}, {tauValues[tauValues.Length - 1]:F2}]");
|
||
}
|
||
#endregion
|
||
|
||
#region 内部类
|
||
private class ValidationResult
|
||
{
|
||
public bool IsValid { get; set; }
|
||
public List<string> Messages { get; } = new List<string>();
|
||
}
|
||
#endregion
|
||
|
||
#region 土壤测试辅助方法（但不影响核心计算）
|
||
/// <summary>
|
||
/// 用于土壤测试的参数设置建议（仅建议，不影响计算）
|
||
/// </summary>
|
||
public static class SoilTestRecommendations
|
||
{
|
||
/// <summary>
|
||
/// 根据土壤类型获取建议的测试参数
|
||
/// </summary>
|
||
/// <param name="soilType">土壤类型（干土、湿土、冻土等）</param>
|
||
/// <returns>建议的测试时间范围</returns>
|
||
public static (double minTime, double maxTime) GetRecommendedTestTime(string soilType)
|
||
{
|
||
// 土壤热扩散系数典型范围：1e-7 到 1e-6 m²/s
|
||
// 使用典型探头半径6.4mm计算
|
||
double r = 0.0064; // 米
|
||
|
||
// 根据τ_max² = (t * a) / r²，反推时间
|
||
// t = τ_max² * r² / a
|
||
|
||
// 对于导热系数较低的土壤（干土、冻土）
|
||
if (soilType.Contains("冻土") || soilType.Contains("干土"))
|
||
{
|
||
double alpha = 0.5e-6; // 较低的热扩散系数
|
||
double t_min = TAU_MAX_SQUARED_LOWER * r * r / alpha;
|
||
double t_max = TAU_MAX_SQUARED_UPPER * r * r / alpha;
|
||
return (t_min, t_max);
|
||
}
|
||
// 对于导热系数较高的土壤（湿土）
|
||
else if (soilType.Contains("湿土"))
|
||
{
|
||
double alpha = 1.2e-6; // 较高的热扩散系数
|
||
double t_min = TAU_MAX_SQUARED_LOWER * r * r / alpha;
|
||
double t_max = TAU_MAX_SQUARED_UPPER * r * r / alpha;
|
||
return (t_min, t_max);
|
||
}
|
||
// 默认值
|
||
else
|
||
{
|
||
double alpha = 1e-6; // 典型热扩散系数
|
||
double t_min = TAU_MAX_SQUARED_LOWER * r * r / alpha;
|
||
double t_max = TAU_MAX_SQUARED_UPPER * r * r / alpha;
|
||
return (t_min, t_max);
|
||
}
|
||
}
|
||
}
|
||
#endregion
|
||
}
|
||
} |