中文说明：应用背景 交互多模型Matlab实现，包含左转，右转和常速度模型。采用Kalman滤波预测估计。效果很好，带有多种误差评价方法及显示结果。 适用目标跟踪，状态空间为位置速度 包含：运动目标跟踪，模型转移概率更新，卡尔曼滤波等文件 关键技术% 基于IMM算法的目标跟踪 % %=============================== %建立模型 %=============================== % 仿真参数 simTime=100; %仿真迭代次数 T=1; %采样时间 w2=3*2*pi/360; %模型2转弯率3度 w3=-3*2*pi/360; %模型3转弯率-3度 H=[1,0,0,0;0,0,1,0]; %模型量测矩阵 G=[T^2/2,0;T,0;0,T^2/2;0,T]; %模型过程噪声加权矩阵 r=200; %20 2000 R=[r,0;0,r]; %模型量测噪声协方差矩阵 Q=[10,0;0,10]; %模型过程噪声协方差矩阵 F1=[1,T,0,0;0,1,0,0;0,0,1,T;0,0,0,1]; %模型1状态转移矩阵 F2=[1,sin(w2*T)/w2,0,(cos(w2*T)-1)/w2; 0,cos(w2*T),0,sin(w2*T); 0,(1-cos(w2*T))/w2,1,sin(w2*T)/w2; 0,-sin(w2*T),0,cos(w2*T)]; %模型2状态转移矩阵 左转弯 F3=[1,sin(w3*T)/w3,0,(cos(w3*T)-1)/w3; 0,cos(w3*T),0,sin(w3
English Description：
Application backgroundInteractive multi model Matlab implementation, including left, right and constant speed model. Prediction and estimation using Kalman filter. Effect is very good, with a variety of error evaluation method and display results.For target tracking, the state space is the position speed.Including: moving target tracking, model transfer probability update, Calman filter and other documentsKey Technology% of target tracking based on IMM algorithm%===============================% model%===============================% of simulation parametersSimTime=100;% simulation iteration numberT=1;% sampling timeW2=3*2*pi/360;% model 2 degree turn rate 3 degreesW3=-3*2*pi/360;% model 3 turning rate -3H=[1,0,0,0; 0,0,1,0]; percent model of measurement matrixG=[T^2/2,0; T, 0; 0; T^2/2; 0; T];% model process noise weighting matrix%20; r=200 2000R=[r, 0; 0; r];% model of measurement noise covariance matrixQ=[10,0; 0,10];% model process noise co