我正在尝试使用 CodeWarrior 在 Freescale Kinetis K70F120M 开发板上运行模糊推理系统 (FIS) 应用程序。
我编写了一个解释器软件，它读取两个纯文本文件（一个包含模糊模型，另一个包含要识别的模式）并编写一个 C++ 应用程序，该应用程序可以上传到开发板以根据传感器读取的数据进行操作。
FIS C++ 应用程序中包含的所有信息都是由解释器编写的，我不可能错误地计算向量的维度，因为元素的数量是从文件中包含的数据中计算出来的。
我已经设法在板上运行了 FIS 示例应用程序，但是当我尝试运行需要运行的实际应用程序时，出现以下 71 个错误：
说明资源路径位置类型
C:\Users\CRISTH~1\AppData\Local\Temp\ccLOvcxh.s 协处理器偏移超出范围
Prueba FALLAS 2
第 8696 行，外部位置：C:\Users\CRISTH~1\AppData\ Local\Temp\ccLOvcxh.s
C/C++ 问题
C:\Users\CRISTH~1\AppData\Local\Temp\ccLOvcxh.s co-processor offset out of range
Prueba FALLAS 2
line 8697，外部位置：C:\Users\ CRISTH~1\AppData\Local\Temp\ccLOvcxh.s
C/C++ 问题
...
C:\Users\CRISTH~1\AppData\Local\Temp\ccLOvcxh.s 协处理器偏移超出范围
Prueba FALLAS 2
行15897 ，外部位置：C:\Users\CRISTH~1\AppData\Local\Temp\ccLOvcxh.s
C/C++ 问题
mingw32-make: *** [Sources/main.o] 错误 1
Prueba FALLAS 2
C/C++ 问题
这是我试图运行的代码：
#include "derivative.h"
#include "network.h"
int main() {
Network ReporteDeFallasCEC;
const int NInputs = 50, NClasses = 10, NRules = 32;
ReporteDeFallasCEC.initialize();
ReporteDeFallasCEC.setClassNeurons(NClasses);
ReporteDeFallasCEC.setRuleNeurons(NRules, 0, Network::Pruning());
Universe input1;
input1.setLimits(0, 100);
input1.addFuzzySet(FuzzySet("vs", 0, 17.52, 35.04));
input1.addFuzzySet(FuzzySet("s", 9.3, 29.65, 50));
input1.addFuzzySet(FuzzySet("m", 23.82, 45.24, 66.67));
input1.addFuzzySet(FuzzySet("l", 50, 66.67, 83.33));
input1.addFuzzySet(FuzzySet("vl", 66.67, 83.33, 100));
ReporteDeFallasCEC.addVariable(input1);
Universe input2;
input2.setLimits(0, 100);
input2.addFuzzySet(FuzzySet("vs", 0, 24.75, 49.51));
input2.addFuzzySet(FuzzySet("s", 5.45, 27.72, 50));
input2.addFuzzySet(FuzzySet("m", 33.33, 5, 6.67));
input2.addFuzzySet(FuzzySet("l", 50, 66.67, 83.33));
input2.addFuzzySet(FuzzySet("vl", 66.67, 83.33, 100));
ReporteDeFallasCEC.addVariable(input2);
//...here goes the remaining "universe" objects
Universe input50;
input50.setLimits(0, 100);
input50.addFuzzySet(FuzzySet("vs", 0, 22.79, 45.57));
input50.addFuzzySet(FuzzySet("s", 0, 34.55, 100));
input50.addFuzzySet(FuzzySet("m", 20.59, 43.63, 66.67));
input50.addFuzzySet(FuzzySet("l", 42.4, 68.49, 95.99));
input50.addFuzzySet(FuzzySet("vl", 66.67, 83.33, 100));
ReporteDeFallasCEC.addVariable(input50);
ClassNeuron InterferenciaDeGas;
ClassNeuron TuberiaDesancladaConGolpeDeFluido;
ClassNeuron RoturaDeVarilla;
ClassNeuron FugaEnLaValvulaFijaDePie;
ClassNeuron FugaEnLaValvulaViajera;
ClassNeuron BarrilDeLaBombaDoblado;
ClassNeuron BuenLlenadoConTuberiaAnclada;
ClassNeuron AgujeroEnElBarrilDeLaBomba;
ClassNeuron AnclaDeTuberiaEnMalFuncionamiento;
ClassNeuron BarrilDeLaBombaGastado;
RuleNeuron FLRule65248697 = RuleNeuron();
FLRule65248697.setNumInputsNeurons(NInputs);
FLRule65248697.setNumClasses(NClasses);
FLRule65248697.setAntecedent(0, 0);
FLRule65248697.setAntecedent(1, 0);
FLRule65248697.setAntecedent(2, 0);
FLRule65248697.setAntecedent(3, 0);
FLRule65248697.setAntecedent(4, 0);
FLRule65248697.setAntecedent(5, 0);
FLRule65248697.setAntecedent(6, 0);
FLRule65248697.setAntecedent(7, 0);
FLRule65248697.setAntecedent(8, 0);
FLRule65248697.setAntecedent(9, 0);
FLRule65248697.setAntecedent(10, 0);
FLRule65248697.setAntecedent(11, 2);
FLRule65248697.setAntecedent(12, 2);
FLRule65248697.setAntecedent(13, 3);
FLRule65248697.setAntecedent(14, 3);
FLRule65248697.setAntecedent(15, 3);
FLRule65248697.setAntecedent(16, 4);
FLRule65248697.setAntecedent(17, 4);
FLRule65248697.setAntecedent(18, 4);
FLRule65248697.setAntecedent(19, 4);
FLRule65248697.setAntecedent(20, 4);
FLRule65248697.setAntecedent(21, 4);
FLRule65248697.setAntecedent(22, 4);
FLRule65248697.setAntecedent(23, 4);
FLRule65248697.setAntecedent(24, 4);
FLRule65248697.setAntecedent(25, 3);
FLRule65248697.setAntecedent(26, 4);
FLRule65248697.setAntecedent(27, 4);
FLRule65248697.setAntecedent(28, 4);
FLRule65248697.setAntecedent(29, 4);
FLRule65248697.setAntecedent(30, 4);
FLRule65248697.setAntecedent(31, 4);
FLRule65248697.setAntecedent(32, 4);
FLRule65248697.setAntecedent(33, 4);
FLRule65248697.setAntecedent(34, 4);
FLRule65248697.setAntecedent(35, 4);
FLRule65248697.setAntecedent(36, 4);
FLRule65248697.setAntecedent(37, 4);
FLRule65248697.setAntecedent(38, 4);
FLRule65248697.setAntecedent(39, 4);
FLRule65248697.setAntecedent(40, 4);
FLRule65248697.setAntecedent(41, 4);
FLRule65248697.setAntecedent(42, 4);
FLRule65248697.setAntecedent(43, 4);
FLRule65248697.setAntecedent(44, 4);
FLRule65248697.setAntecedent(45, 4);
FLRule65248697.setAntecedent(46, 4);
FLRule65248697.setAntecedent(47, 4);
FLRule65248697.setAntecedent(48, 4);
FLRule65248697.setAntecedent(49, 4);
FLRule65248697.setConsecuent(1);
ReporteDeFallasCEC.addRule(FLRule65248697);
RuleNeuron FLRule50510248 = RuleNeuron();
FLRule50510248.setNumInputsNeurons(NInputs);
FLRule50510248.setNumClasses(NClasses);
FLRule50510248.setAntecedent(0, 0);
FLRule50510248.setAntecedent(1, 0);
FLRule50510248.setAntecedent(2, 0);
FLRule50510248.setAntecedent(3, 0);
FLRule50510248.setAntecedent(4, 0);
FLRule50510248.setAntecedent(5, 0);
FLRule50510248.setAntecedent(6, 0);
FLRule50510248.setAntecedent(7, 0);
FLRule50510248.setAntecedent(8, 0);
FLRule50510248.setAntecedent(9, 0);
FLRule50510248.setAntecedent(10, 0);
FLRule50510248.setAntecedent(11, 0);
FLRule50510248.setAntecedent(12, 0);
FLRule50510248.setAntecedent(13, 0);
FLRule50510248.setAntecedent(14, 2);
FLRule50510248.setAntecedent(15, 4);
FLRule50510248.setAntecedent(16, 4);
FLRule50510248.setAntecedent(17, 4);
FLRule50510248.setAntecedent(18, 4);
FLRule50510248.setAntecedent(19, 4);
FLRule50510248.setAntecedent(20, 4);
FLRule50510248.setAntecedent(21, 4);
FLRule50510248.setAntecedent(22, 4);
FLRule50510248.setAntecedent(23, 4);
FLRule50510248.setAntecedent(24, 4);
FLRule50510248.setAntecedent(25, 3);
FLRule50510248.setAntecedent(26, 4);
FLRule50510248.setAntecedent(27, 4);
FLRule50510248.setAntecedent(28, 4);
FLRule50510248.setAntecedent(29, 4);
FLRule50510248.setAntecedent(30, 4);
FLRule50510248.setAntecedent(31, 4);
FLRule50510248.setAntecedent(32, 4);
FLRule50510248.setAntecedent(33, 4);
FLRule50510248.setAntecedent(34, 4);
FLRule50510248.setAntecedent(35, 4);
FLRule50510248.setAntecedent(36, 4);
FLRule50510248.setAntecedent(37, 4);
FLRule50510248.setAntecedent(38, 4);
FLRule50510248.setAntecedent(39, 4);
FLRule50510248.setAntecedent(40, 4);
FLRule50510248.setAntecedent(41, 4);
FLRule50510248.setAntecedent(42, 4);
FLRule50510248.setAntecedent(43, 4);
FLRule50510248.setAntecedent(44, 4);
FLRule50510248.setAntecedent(45, 4);
FLRule50510248.setAntecedent(46, 4);
FLRule50510248.setAntecedent(47, 4);
FLRule50510248.setAntecedent(48, 4);
FLRule50510248.setAntecedent(49, 4);
FLRule50510248.setConsecuent(2);
ReporteDeFallasCEC.addRule(FLRule50510248);
//...here goes the remaining "RuleNeuron" objects
RuleNeuron FLRule2056998 = RuleNeuron();
FLRule2056998.setNumInputsNeurons(NInputs);
FLRule2056998.setNumClasses(NClasses);
FLRule2056998.setAntecedent(0, 0);
FLRule2056998.setAntecedent(1, 0);
FLRule2056998.setAntecedent(2, 0);
FLRule2056998.setAntecedent(3, 0);
FLRule2056998.setAntecedent(4, 0);
FLRule2056998.setAntecedent(5, 0);
FLRule2056998.setAntecedent(6, 0);
FLRule2056998.setAntecedent(7, 0);
FLRule2056998.setAntecedent(8, 0);
FLRule2056998.setAntecedent(9, 0);
FLRule2056998.setAntecedent(10, 0);
FLRule2056998.setAntecedent(11, 1);
FLRule2056998.setAntecedent(12, 1);
FLRule2056998.setAntecedent(13, 1);
FLRule2056998.setAntecedent(14, 1);
FLRule2056998.setAntecedent(15, 2);
FLRule2056998.setAntecedent(16, 2);
FLRule2056998.setAntecedent(17, 3);
FLRule2056998.setAntecedent(18, 3);
FLRule2056998.setAntecedent(19, 4);
FLRule2056998.setAntecedent(20, 3);
FLRule2056998.setAntecedent(21, 4);
FLRule2056998.setAntecedent(22, 3);
FLRule2056998.setAntecedent(23, 3);
FLRule2056998.setAntecedent(24, 4);
FLRule2056998.setAntecedent(25, 0);
FLRule2056998.setAntecedent(26, 0);
FLRule2056998.setAntecedent(27, 0);
FLRule2056998.setAntecedent(28, 0);
FLRule2056998.setAntecedent(29, 0);
FLRule2056998.setAntecedent(30, 0);
FLRule2056998.setAntecedent(31, 0);
FLRule2056998.setAntecedent(32, 0);
FLRule2056998.setAntecedent(33, 1);
FLRule2056998.setAntecedent(34, 1);
FLRule2056998.setAntecedent(35, 1);
FLRule2056998.setAntecedent(36, 1);
FLRule2056998.setAntecedent(37, 2);
FLRule2056998.setAntecedent(38, 2);
FLRule2056998.setAntecedent(39, 2);
FLRule2056998.setAntecedent(40, 2);
FLRule2056998.setAntecedent(41, 3);
FLRule2056998.setAntecedent(42, 3);
FLRule2056998.setAntecedent(43, 3);
FLRule2056998.setAntecedent(44, 4);
FLRule2056998.setAntecedent(45, 4);
FLRule2056998.setAntecedent(46, 4);
FLRule2056998.setAntecedent(47, 4);
FLRule2056998.setAntecedent(48, 3);
FLRule2056998.setAntecedent(49, 4);
FLRule2056998.setConsecuent(3);
ReporteDeFallasCEC.addRule(FLRule2056998);
const int nPatterns = 60;
double patternArray[nPatterns][NInputs] = {
{ 6.60, 9.70, 12.2, 4.60, 5.70, 8.70, 12.9, 7.60, 11.8, 7.90, 24.9, 44.3, 55.0, 63.6, 71.3, 75.0, 76.8, 80.9, 84.50, 86.50, 90.70, 91.40, 95.00, 97.30, 93.80, 69.20, 82.60, 95.80, 98.90, 97.20, 97.70, 97.80, 96.90, 97.50, 94.50, 93.50, 95.80, 92.50, 93.60, 94.20, 92.00, 90.20, 91.60, 90.20, 91.50, 91.80, 90.70, 93.90, 96.10, 95.70 },
{ 7.70, 5.50, 4.50, 0.60, 1.70, 5.90, 6.70, 6.70, 8.60, 10.1, 5.60, 5.30, 8.40, 24.3, 57.2, 79.8, 88.0, 90.8, 91.10, 91.90, 92.40, 91.80, 91.60, 91.40, 95.70, 62.20, 94.10, 96.40, 91.60, 92.40, 94.00, 97.40, 97.70, 97.80, 96.10, 94.40, 94.40, 96.00, 98.10, 99.30, 94.40, 94.50, 96.30, 96.80, 94.20, 94.50, 96.60, 98.70, 97.10, 97.30 },
//...here goes the remaining pattern rows
{ 1.95, 3.10, 2.00, 3.20, 1.90, 3.70, 5.70, 5.00, 2.10, 3.50, 3.30, 2.05, 2.95, 2.40, 2.70, 3.80, 4.10, 3.30, 4.700, 5.100, 5.300, 5.900, 8.600, 14.30, 38.70, 87.40, 94.80, 96.00, 97.00, 97.40, 98.40, 98.80, 98.10, 98.90, 96.50, 92.00, 83.80, 75.00, 67.80, 69.00, 77.00, 87.70, 94.90, 96.00, 95.80, 95.60, 95.70, 93.80, 87.60, 38.80 }, };
double pattern[NInputs];
for (int i = 0; i < nPatterns; i++) {
std::cout << "Patron " << i + 1 << ": [ ";
for (int j = 0; j < NInputs; j++) {
pattern[j] = patternArray[i][j];
std::cout << patternArray[i][j];
if (j != NInputs - 1) {
std::cout << ", ";
}
}
std::cout << " ]" << "\n";
int clase = ReporteDeFallasCEC.classificate(pattern);
std::cout << "\nClase [ " << clase << " ]\n\n";
}
//std::cin.get();
return 0;
}
此 CodeWarrior 项目是在 CodeWarrior IDE v.10.6.4 中作为新的裸板项目创建的：
要使用的设备：MK70FN1M0 处理器（K70F 120 MHz 系列）
项目类型：
要使用的应用程序连接：开源 JTAG
语言：C++
浮点：硬件 (-mfloat-abi=hard) 与 (-fp vfpv4)
I/O 支持：调试器控制台
ARM 构建工具：GCC
快速应用程序开发：无
从设计的透视图开始：当前透视图
我正在使用 FLASH 配置和调试作为“Prueba FALLAS 2_FLASH_OSJTAG”来构建项目。
我在 Windows 7 上运行 CodeWarrior。
帮我找出阻止代码在板上运行的原因。
更新 #1： 我已经删除了关于“ClassNeuron”和“RuleNeuron”对象的生成以及 patternArray 向量的代码，包括使用该向量的函数，所以剩下的应用程序只创建了 50 个“Universe”对象. 这样做之后，我开始删除故意数量的 Universe 对象，以尝试找出是否存在与内存限制相关的问题，但无论是 15 个还是 26 个对象，我都会得到上述类型的随机数量的错误（甚至0 个错误）。我需要使用确切数量的对象（50 个 Universe、10 个类、32 个规则和 50x60 模式阵列）运行我的应用程序。
我怀疑这个问题与代码创建的对象数量有关，但我不确定在编译项目时 CodeWarrior 是否设置了 FLASH 内存限制。尽管如此，我很确定这些错误与数组处理无关，因为我已经从项目中彻底删除了对 patternArray 的任何引用，而且，如果我对数组做错了什么，CodeWarrior IDE 应该给出我对此有一些线索。
帮助我解决这个问题，因为即使在 NXP（飞思卡尔）社区中也很难找到有关此问题的信息。
更新＃2： 如以下相关问题所述，看来这个错误实际上是编译器错误，请向我确认：
这是我在 NXP（原飞思卡尔）社区的帖子：协处理器偏移超出范围
部分解决
我将我的代码移到了 Kinetis Design Studio 中的一个项目中。它编译没有错误，我可以调试应用程序，直到电路板内存不足的某个时候，然后我应用了一些优化更改，一切都很顺利。