def RadarDataRelateBasedonMachineLearning(realData = False, machineLearnEnable = False,mlModelType = 'KNN', totalSimuFile = 30, totalFramePerSimu = 10):
mlModel = None
if machineLearnEnable == True:#训练机器学习模型
mlModel = RadarMl.RadarML(mlModelType)
mlModel.DatasetProc(simu=True,scale_en= True, poly_en=False)
mlModel.Train()
if realData == True:#如果利用真实航迹作为数据源，则需要将真实数据进行分帧处理
Frame.FrameCreat(save_en=True, plot_en=True, fakerate=50, sample_rate=5)
for simuFileNo in range(totalSimuFile): #对每一个仿真文件进行航迹关联处理
processOutput = sys.stdout # 标准图像输出
plt.figure(figsize=(5, 40), dpi=100) # 设置输出图像画布大小
plt.vlines(x=[-5.5, -1.8, 1.8, 5.5], ymin=0, ymax=40, colors='yellow') # 画出车道线
tmpTracksList, totalTmpTracks = tmpTracksListInit() #初始化临时航迹列表
tmpStoragePoints,totalTmpStoragePoints = tmpStoragePointsInit() #初始化临时点迹列表
for frameNo in range(totalFramePerSimu): #从文件中读取每帧量测
if realData == True:
frameInfor = Frame.FrameRead(frameNo)
else:
frameInfor = Frame.SimuFrameRead(simuFileNo,frameNo)
count = frameNo / (totalFramePerSimu - 1) * 100 #计算显示每个仿真文件的处理进程
processOutput.write(f'\r PROCESSING percent:{count:.0f}%')
if machineLearnEnable == True:#机器学习模型应用于雷达量测分类
frameInfor, mlMissPoints = mlModel.Applicate(frameInfor,scale_en=True)
else:
mlMissPoints = pd.DataFrame([],columns=['Angle','Speed','X_position','Y_position','Target'])
#对雷达量测进行航迹关联
[frameInfor, tmpTracksList, tmpStoragePoints, totalTmpStoragePoints] = TrackRelate(frameInfor, tmpTracksList, totalTmpTracks, tmpStoragePoints, totalTmpStoragePoints,mlModel,machineLearnEnable = machineLearnEnable)
#删除关联失败的航迹信息
[tmpTracksList, totalTmpTracks, tmpStoragePoints, totalTmpStoragePoints, delete_points] = TrackDelet(tmpTracksList, totalTmpTracks, tmpStoragePoints, totalTmpStoragePoints)
#没有正确关联的点作为新航迹
[tmpTracksList, totalTmpTracks] = TrackNewDevelop(frameInfor, tmpTracksList, totalTmpTracks)
#绘制关联成功的航迹
[tmpStoragePoints, totalTmpStoragePoints] = TrackDraw(frameInfor, tmpTracksList, tmpStoragePoints, totalTmpStoragePoints, delete_points,mlMissPoints=mlMissPoints)
processOutput.flush
plt.xlim((-7, 7))#坐标属性设置
plt.ylim((0, 40))
plt.xlabel("X(m)")
plt.ylabel("Y(m)")
plt.show()
plt.pause(0)#暂停显示关联结果
return
