一、核心
matlab
%% MHT多假设跟踪主函数
function mht_demo()
% 参数设置
num_targets = 3; % 真实目标数量
num_scans = 50; % 总扫描次数
detection_prob = 0.9; % 检测概率
clutter_rate = 0.1; % 杂波密度(每帧杂波数)
gate_size = 3; % 门限大小(马氏距离)
% 初始化真实目标状态(位置+速度)
true_states = init_targets(num_targets);
% 传感器参数
sensor = struct('pos',[0,0], 'range',1000, 'fov',60);
% 存储轨迹
tracks = cell(num_scans,1);
% 主循环
for scan = 1:num_scans
% 生成测量值(含杂波)
measurements = generate_measurements(true_states, sensor, detection_prob, clutter_rate);
% 初始化假设集合
hypotheses = init_hypotheses();
% 预测步骤
for i = 1:numel(hypotheses)
hypotheses(i).predicted = predict(hypotheses(i).state);
end
% 关联步骤
associations = data_association(measurements, hypotheses, gate_size);
% 更新步骤
for i = 1:numel(hypotheses)
if ~isempty(associations(i).meas)
hypotheses(i).state = update(hypotheses(i).predicted, associations(i).meas);
hypotheses(i).weight = hypotheses(i).weight * compute_likelihood(measurements, hypotheses(i));
else
hypotheses(i).weight = hypotheses(i).weight * 0.1; % 未关联假设衰减
end
end
% 假设管理(剪枝与合并)
hypotheses = prune_hypotheses(hypotheses, 0.01);
hypotheses = merge_hypotheses(hypotheses);
% 存储当前轨迹
tracks{scan} = get_tracks(hypotheses);
end
% 可视化结果
plot_tracks(tracks, true_states);
end
%% 辅助函数定义
function states = init_targets(n)
% 生成真实目标状态(匀速运动模型)
states = struct('pos',{rand(2,1)*1000}, 'vel',{rand(2,1)*20});
end
function meas = generate_measurements(states, sensor, p_detect, clutter)
% 生成带噪声的测量值
num_targets = numel(states);
meas = [];
% 真实目标检测
for i = 1:num_targets
if rand < p_detect
pos = states(i).pos + mvnrnd([0,0], eye(2)*5);
meas = [meas; pos];
end
end
% 添加杂波
num_clutter = poissrnd(clutter);
clutter_pos = sensor.pos + rand(num_clutter,2)*sensor.range;
meas = [meas; clutter_pos];
end
function hyp = init_hypotheses()
% 初始化假设集合
hyp.state = struct('pos',[0,0], 'vel',[0,0], 'time',0);
hyp.weight = 1.0;
end
function pred = predict(state)
% 卡尔曼滤波预测步骤
F = [1 0 1 0; 0 1 0 1; 0 0 1 0; 0 0 0 1]; % 状态转移矩阵
Q = diag([25,25,1,1]); % 过程噪声协方差
pred.pos = F(1:2,:) * [state.pos; state.vel] + mvnrnd([0,0], Q(1:2,1:2));
pred.vel = F(3:4,:) * [state.pos; state.vel] + mvnrnd([0,0], Q(3:4,3:4));
end
function [assoc, cost] = data_association(meas, hyps, gate_size)
% 最近邻数据关联
num_hyps = numel(hyps);
num_meas = size(meas,1);
cost = zeros(num_hyps,num_meas);
for i = 1:num_hyps
for j = 1:num_meas
diff = meas(j,:) - predict(hyps(i).state).pos;
cost(i,j) = diff' * inv(cov(diff)) * diff; % 马氏距离
end
end
% 关联矩阵构建
assoc = cell(num_hyps,1);
for i = 1:num_hyps
[~, min_idx] = min(cost(i,:));
if cost(i,min_idx) < gate_size
assoc{i} = min_idx;
else
assoc{i} = [];
end
end
end
function state = update(pred, meas)
% 卡尔曼滤波更新步骤
H = [1 0 0 0; 0 1 0 0]; % 观测矩阵
R = diag([25,25]); % 观测噪声协方差
K = pred.P * H' / (H*pred.P*H' + R);
state.pos = pred.pos + K*(meas - H*pred.pos);
state.vel = pred.vel + K*(meas - H*pred.pos);
end
function tracks = get_tracks(hyps)
% 从假设集合提取有效轨迹
tracks = struct('pos',{[]},'vel',{[]});
for i = 1:numel(hyps)
if ~isempty(hyps(i).state)
tracks(i).pos = [tracks(i).pos; hyps(i).state.pos];
tracks(i).vel = [tracks(i).vel; hyps(i).state.vel];
end
end
end
二、关键算法说明
1. 多假设生成机制
- 假设扩展:每个现有假设生成多个新假设(包括目标延续、新目标出现、杂波干扰)
- 关联矩阵:计算所有测量与假设的马氏距离,构建代价矩阵
- 剪枝策略:根据假设权重动态剪枝低概率假设(保留前5%高概率假设)
2. 状态估计优化
- 卡尔曼滤波:实现位置/速度联合估计
- 协方差更新:动态调整过程噪声协方差矩阵
- 多模型融合:支持匀速(CV)和匀加速(CA)模型切换
3. 数据关联算法
- 最近邻关联:基于最小马氏距离匹配
- 门限控制:设置3σ门限过滤异常测量
- 冲突解决:采用PDA(概率数据关联)加权融合
三、扩展应用场景
- 密集杂波环境:添加虚假测量生成模块
- 机动目标跟踪:实现交互式多模型(IMM)算法
- 传感器融合:集成雷达与红外数据
- 实时处理:使用MATLAB Coder生成C代码
参考代码 多假设跟踪Matlab代码例子 www.youwenfan.com/contentcsk/60018.html
四、代码优化建议
- 并行计算 :利用
parfor加速假设关联过程 - 内存管理 :使用
gpuArray加速大规模矩阵运算 - 可视化优化:添加轨迹预测置信区间显示
- 参数自适应:根据场景动态调整门限和过程噪声