✅作者简介热爱科研的Matlab仿真开发者擅长毕业设计辅导、数学建模、数据处理、建模仿真、程序设计、完整代码获取、论文复现及科研仿真。 往期回顾关注个人主页Matlab科研工作室 关注我领取海量matlab电子书和数学建模资料个人信条做科研博学之、审问之、慎思之、明辨之、笃行之是为博学慎思明辨笃行。 内容介绍这份 MATLAB 代码是城市迷宫场景下多无人机自适应编队协同避障仿真框架实现环境感知熵 障碍物密度双指标自动切换编队、V 型初始编队、领航者 - 跟随者分层控制、碰撞检测、卡死检测、指标绘图、数据保存、视频输出整套流水线。代码为越南语注释我先逐段翻译讲解逻辑再给出全中文注释优化版完整代码。一、整体架构分层6 大核心模块初始化模块随机种子、场景选择、环境建模、编队选择器初始化无人机初始化V 型编队布放领航机 / 跟随机根据起点 - 目标航向计算初始偏移仿真全局参数配置仿真步长、最大迭代、熵归一化阈值、碰撞 / 卡死计数器主仿真循环核心全局计算编队熵、障碍物密度、编队类型自适应切换带滞环防频繁跳变单机更新领航者避障寻目标、跟随者跟踪编队几何、碰撞逃逸力、速度限幅故障检测机身碰撞计数、领航机卡死判定告警数据缓存熵、编队类型、障碍物密度时序记录后处理分析仿真终止判定、碰撞 / 编队切换次数统计、性能指标打印可视化与存储静态结果绘图、性能指标评估、仿真视频生成、mat 数据保存二、关键核心算法说明1. 自适应编队切换逻辑两个感知特征作为输入entropy编队熵无人机分布混乱度熵越大编队越散乱需要收缩紧密编队rho障碍物密度领航机周边障碍物占比密度高切换窄通道紧凑型编队FormationSelector内置滞环缓存hysteresis_buffer、最小切换间隔min_switch_interval防止场景小幅扰动导致编队频繁抖动城市迷宫场景专用参数。2. V 型编队几何生成3 号无人机固定为领航 Leader跟随机按距离领航机距离做径向偏移左侧机翼航向 夹角 α180°置于领航机后方右侧机翼航向 - 夹角 α180°保证编队整体朝向目标点初始队形标准 V 型。3. 分层运动控制领航机 LeaderBehavior全局导航向目标 全局避障主导编队前进方向跟随机 FollowerBehavior仅跟踪编队几何约束跟随领航机轨迹碰撞逃逸机制无人机落入多边形障碍物时叠加法向斥力 8 单位强制脱离速度上限 2m/s 限幅。4. 故障监测机制碰撞计数任意无人机进入障碍物多边形即计数领航机卡死检测连续 200 迭代位移小于 0.05m 判定卡死并打印告警5. 场景熵归一化适配不同环境最大混乱熵阈值不同归一化到 [0,1] 区间保证编队选择器输入尺度统一窄通道20.0 | 城市迷宫5.0 | 开阔场地15.0⛳️ 运行结果 部分代码function PlotResultsSmart(drones, model, selector, history_entropy, history_formation, history_rho)%PLOTRESULTSSMART - Vẽ kết quả mô phỏng với thông tin đội hình thông minh%% Tham sốnum_steps length(history_entropy);if num_steps 0warning(No history data to plot);return;end% Kiểm tra consistencypath_steps size(drones{1}.path, 1);if path_steps ~ num_steps 1warning(Path steps (%d) ~ history steps (%d) 1. Using min., ...path_steps, num_steps);num_steps min(num_steps, path_steps - 1);enddt 0.02;% Màu theo vai tròrole_colors struct();role_colors.leader [1.0, 0.0, 0.0];role_colors.left_wing [0.0, 0.4, 1.0];role_colors.right_wing [0.0, 0.8, 0.0];role_colors.tail [1.0, 0.8, 0.0];role_colors.center [0.8, 0.0, 0.8];role_colors.perimeter [0.5, 0.5, 0.5];role_colors.left_guard [0.0, 0.6, 0.8];role_colors.right_guard [0.8, 0.4, 0.0];role_colors.wing [0.2, 0.6, 0.2];role_colors.front [0.4, 0.4, 0.8];role_colors.rear [0.6, 0.6, 0.6];role_colors.reserve [0.6, 0.6, 0.6];role_colors.follower [0.4, 0.4, 0.8];default_color [0.3, 0.3, 0.3];% Tính transition_pointstransition_points [1];transition_types {history_formation{1}};for i 2:length(history_formation)if ~strcmp(history_formation{i}, history_formation{i-1})transition_points [transition_points, i];transition_types{end1} history_formation{i};endend%% Tạo figurefig figure(Name, sprintf(Smart Formation Results - %s, model.scenario), ...Position, [30, 30, 1500, 1000], ...Color, [1, 1, 1]);%% SUBPLOT 1: Quỹ đạo tổng thể (lớn) ax1 subplot(3, 3, [1, 2, 4, 5]);hold(ax1, on);grid(ax1, on);% Môi trườngplot(ax1, model.start(1), model.start(2), bs, ...MarkerSize, 20, MarkerFaceColor, [0.2, 0.2, 1.0], ...MarkerEdgeColor, k, LineWidth, 2.5, ...DisplayName, Start);text(ax1, model.start(1), model.start(2)-1.0, START, ...HorizontalAlignment, center, FontSize, 12, ...FontWeight, bold, Color, [0.2, 0.2, 1.0]);plot(ax1, model.goal(1), model.goal(2), rp, ...MarkerSize, 20, MarkerFaceColor, [1.0, 0.2, 0.2], ...MarkerEdgeColor, k, LineWidth, 2.5, ...DisplayName, Goal);text(ax1, model.goal(1), model.goal(2)-1.0, GOAL, ...HorizontalAlignment, center, FontSize, 12, ...FontWeight, bold, Color, [1.0, 0.2, 0.2]);% Obstaclesfor j 1:size(model.obstacles, 2)obs model.obstacles{j};pgon polyshape(obs(:, 1), obs(:, 2));plot(ax1, pgon, FaceColor, [0.25, 0.25, 0.25], ...FaceAlpha, 0.7, EdgeColor, k, LineWidth, 2, ...DisplayName, sprintf(Obstacle %d, j));end% Quỹ đạo UAVfor i 1:model.nrole drones{i}.formation_role;if isfield(role_colors, role)color role_colors.(role);elsecolor default_color;endpath drones{i}.path;plot(ax1, path(:, 1), path(:, 2), ...Color, color, LineWidth, 2.5, ...DisplayName, sprintf(UAV%d (%s), i, role));end% Snapshot đội hìnhsnapshot_points [1];if length(transition_points) 1snapshot_points [snapshot_points, transition_points(2:end)];endif snapshot_points(end) ~ num_stepssnapshot_points [snapshot_points, num_steps];endsnapshot_points unique(snapshot_points);snapshot_points snapshot_points(snapshot_points num_steps);snapshot_colors lines(length(snapshot_points));for s 1:length(snapshot_points)idx snapshot_points(s);if idx length(history_formation)continue;endgr [];for j 1:model.npath_idx min(idx 1, size(drones{j}.path, 1));gr [gr; drones{j}.path(path_idx, 1:2)];end% Vẽ đội hìnhplot(ax1, gr(:, 1), gr(:, 2), --, ...Color, [snapshot_colors(s, :), 0.6], LineWidth, 1.5);scatter(ax1, gr(:, 1), gr(:, 2), 100, ...snapshot_colors(s, :), filled, ...MarkerEdgeColor, k, LineWidth, 1.5);% Label đội hìnhmid_x mean(gr(:, 1));mid_y max(gr(:, 2)) 0.8;text(ax1, mid_x, mid_y, ...sprintf([%s]\nt%.1fs, history_formation{idx}, idx*dt), ...HorizontalAlignment, center, FontSize, 9, ...Color, snapshot_colors(s, :), FontWeight, bold, ...BackgroundColor, [1, 1, 1, 0.8], EdgeColor, snapshot_colors(s, :));endxlabel(ax1, x [m], FontSize, 12, FontWeight, bold);ylabel(ax1, y [m], FontSize, 12, FontWeight, bold);title(ax1, sprintf(UAV Trajectories - %s, model.scenario), ...FontSize, 14, FontWeight, bold);legend(ax1, Location, bestoutside, FontSize, 9);axis(ax1, equal);xlim(ax1, [model.xmin, model.xmax]);ylim(ax1, [model.ymin, model.ymax]);%% SUBPLOT 2: Timeline đội hình ax2 subplot(3, 3, 3);hold(ax2, on);if ~isempty(history_formation)hist_form history_formation(1:num_steps);formation_types unique(hist_form);num_types length(formation_types);colors jet(num_types);% Vẽ timelinefor i 1:num_steps-1type_idx find(strcmp(formation_types, hist_form{i}));fill(ax2, [i, i1, i1, i], [0, 0, 1, 1], ...colors(type_idx, :), FaceAlpha, 0.7, EdgeColor, none);end% Đánh dấu transitionfor i 2:length(transition_points)if transition_points(i) num_stepsline(ax2, [transition_points(i), transition_points(i)], [0, 1], ...Color, k, LineStyle, --, LineWidth, 2);text(ax2, transition_points(i), 1.1, transition_types{i}, ...HorizontalAlignment, center, FontSize, 8, ...Rotation, 45, Color, k, FontWeight, bold);endendyticks(ax2, []);ylim(ax2, [0, 1.5]);xlabel(ax2, Time step, FontSize, 10);title(ax2, Formation Timeline, FontSize, 11, FontWeight, bold);% Legendh_dummy [];for i 1:num_typesh_dummy(i) plot(ax2, nan, nan, s, Color, colors(i, :), ...MarkerFaceColor, colors(i, :), ...MarkerSize, 10, DisplayName, formation_types{i});endlegend(ax2, h_dummy, Location, best, FontSize, 9);end%% SUBPLOT 3: Entropy theo thời gian (ĐÃ SỬA YLIM) ax3 subplot(3, 3, 6);hold(ax3, on);grid(ax3, on);if ~isempty(history_entropy)plot(ax3, 1:num_steps, history_entropy(1:num_steps), b-, LineWidth, 2);% Ngưỡngline(ax3, [1, num_steps], [0.8, 0.8], Color, [1, 0.5, 0], ...LineStyle, --, LineWidth, 1.5, DisplayName, High threshold);line(ax3, [1, num_steps], [0.5, 0.5], Color, [0.8, 0.8, 0], ...LineStyle, --, LineWidth, 1.5, DisplayName, Medium threshold);% Đánh dấu transitionfor i 2:length(transition_points)if transition_points(i) num_stepsxline(ax3, transition_points(i), --g, LineWidth, 1.5);endend% ✅ SỬA: Ylim động dựa trên giá trị entropy thực tếmax_entropy max(history_entropy);y_max max(max_entropy * 1.1, 1.2); % Ít nhất 1.2, hoặc cao hơn nếu cầnxlim(ax3, [1, num_steps]);ylim(ax3, [0, y_max]);endxlabel(ax3, Time step, FontSize, 10);ylabel(ax3, Entropy, FontSize, 10);title(ax3, Formation Entropy, FontSize, 11, FontWeight, bold);legend(ax3, Location, best, FontSize, 9);%% SUBPLOT 4: Obstacle Density ax4 subplot(3, 3, 9);hold(ax4, on);grid(ax4, on);if ~isempty(history_rho)plot(ax4, 1:num_steps, history_rho(1:num_steps), r-, LineWidth, 2);% Ngưỡngline(ax4, [1, num_steps], [0.6, 0.6], Color, [1, 0.3, 0.3], ...LineStyle, --, LineWidth, 1.5, DisplayName, High density);% Đánh dấu transitionfor i 2:length(transition_points)if transition_points(i) num_stepsxline(ax4, transition_points(i), --g, LineWidth, 1.5);endendxlim(ax4, [1, num_steps]);ylim(ax4, [0, 1]);endxlabel(ax4, Time step, FontSize, 10);ylabel(ax4, Obstacle Density \rho, FontSize, 10);title(ax4, Obstacle Density, FontSize, 11, FontWeight, bold);legend(ax4, Location, best, FontSize, 9);%% SUBPLOT 5: Heading Order ax5 subplot(3, 3, 7);hold(ax5, on);grid(ax5, on);headings [];for i 1:num_stepsheading_vec [0, 0];for j 1:model.npath_idx min(i 1, size(drones{j}.path, 1));h drones{j}.path(path_idx, 3);heading_vec heading_vec [cos(h), sin(h)];endheadings [headings, norm(heading_vec) / model.n];endplot(ax5, 1:num_steps, headings, Color, [0.2, 0.6, 0.2], LineWidth, 2);% Đánh dấu transitionfor i 2:length(transition_points)if transition_points(i) num_stepsxline(ax5, transition_points(i), --g, LineWidth, 1.5);endendxlim(ax5, [1, num_steps]);ylim(ax5, [0, 1.1]);xlabel(ax5, Time step, FontSize, 10);ylabel(ax5, Order, FontSize, 10);title(ax5, Heading Consensus, FontSize, 11, FontWeight, bold);%% SUBPLOT 6: Inter-UAV Distance ax6 subplot(3, 3, 8);hold(ax6, on);grid(ax6, on);fill(ax6, [1, num_steps, num_steps, 1], ...[0, 0, drones{1}.ra, drones{1}.ra], ...[1, 0.8, 0.8], FaceAlpha, 0.3, EdgeColor, none, ...DisplayName, Collision Zone);line(ax6, [1, num_steps], [drones{1}.ra, drones{1}.ra], ...Color, r, LineStyle, --, LineWidth, 2, ...DisplayName, Safety threshold);pair_names {};for i 1:model.n-1for j i1:model.npath_i drones{i}.path(2:min(num_steps1, size(drones{i}.path, 1)), 1:2);path_j drones{j}.path(2:min(num_steps1, size(drones{j}.path, 1)), 1:2);min_len min(size(path_i, 1), size(path_j, 1));path_i path_i(1:min_len, :);path_j path_j(1:min_len, :);dis path_i - path_j;dist_vec sqrt(sum(dis.^2, 2));if length(dist_vec) num_stepsdist_vec [dist_vec; repmat(dist_vec(end), num_steps - length(dist_vec), 1)];endpair_name sprintf(UAV%d-UAV%d, i, j);pair_names{end1} pair_name;plot(ax6, 1:num_steps, dist_vec(1:num_steps), LineWidth, 1.2, ...DisplayName, pair_name);endend% Đánh dấu transitionfor i 2:length(transition_points)if transition_points(i) num_stepsxline(ax6, transition_points(i), --g, LineWidth, 1.5);endendxlim(ax6, [1, num_steps]);xlabel(ax6, Time step, FontSize, 10);ylabel(ax6, Distance [m], FontSize, 10);title(ax6, Inter-UAV Distance, FontSize, 11, FontWeight, bold);legend(ax6, NumColumns, 2, Location, best, FontSize, 8);%% Lưu figure sgtitle(fig, sprintf(Smart Formation Control Results - %s, model.scenario), ...FontSize, 16, FontWeight, bold, Color, [0.2, 0.2, 0.2]);filename_png sprintf(result_smart_%s.png, model.scenario);saveas(fig, filename_png);fprintf(Figure saved: %s\n, filename_png);%% In báo cáo tóm tắt fprintf(\n);fprintf(╔══════════════════════════════════════════════════════════╗\n);fprintf(║ SMART FORMATION PLOT REPORT ║\n);fprintf(╠══════════════════════════════════════════════════════════╣\n);fprintf(║ Scenario: %-37s ║\n, model.scenario);fprintf(║ Total time: %-8.2f s ║\n, num_steps * dt);fprintf(║ Total iterations: %-8d ║\n, num_steps);fprintf(║ Number of UAVs: %-8d ║\n, model.n);fprintf(║ Transitions: %-8d ║\n, length(transition_points)-1);fprintf(╠══════════════════════════════════════════════════════════╣\n);fprintf(║ Formation sequence: ║\n);for i 1:length(transition_points)duration 0;if i length(transition_points)duration (transition_points(i1) - transition_points(i)) * dt;elseduration (num_steps - transition_points(i) 1) * dt;endfprintf(║ %-10s t%-6.2fs (duration: %-6.2fs) ║\n, ...transition_types{i}, transition_points(i)*dt, duration);endfprintf(╠══════════════════════════════════════════════════════════╣\n);fprintf(║ Avg Entropy: %-8.4f ║\n, mean(history_entropy));fprintf(║ Max Entropy: %-8.4f ║\n, max(history_entropy));fprintf(║ Avg Obstacle Density: %-8.4f ║\n, mean(history_rho));fprintf(║ Avg Heading Order: %-8.4f ║\n, mean(headings));fprintf(║ Min Heading Order: %-8.4f ║\n, min(headings));fprintf(╚══════════════════════════════════════════════════════════╝\n);end 参考文献更多创新智能优化算法模型和应用场景可扫描关注机器学习/深度学习类BP、SVM、RVM、DBN、LSSVM、ELM、KELM、HKELM、DELM、RELM、DHKELM、RF、SAE、LSTM、BiLSTM、GRU、BiGRU、PNN、CNN、XGBoost、LightGBM、TCN、BiTCN、ESN、Transformer、模糊小波神经网络、宽度学习等等均可~方向涵盖风电预测、光伏预测、电池寿命预测、辐射源识别、交通流预测、负荷预测、股价预测、PM2.5浓度预测、电池健康状态预测、用电量预测、水体光学参数反演、NLOS信号识别、地铁停车精准预测、变压器故障诊断组合预测类CNN/TCN/BiTCN/DBN/Transformer/Adaboost结合SVM、RVM、ELM、LSTM、BiLSTM、GRU、BiGRU、Attention机制类等均可可任意搭配非常新颖~分解类EMD、EEMD、VMD、REMD、FEEMD、TVFEMD、CEEMDAN、ICEEMDAN、SVMD、FMD、JMD等分解模型均可~路径规划类旅行商问题TSP、车辆路径问题VRP、MVRP、CVRP、VRPTW等、无人机三维路径规划、无人机协同、无人机编队、机器人路径规划、栅格地图路径规划、多式联运运输问题、 充电车辆路径规划EVRP、 双层车辆路径规划2E-VRP、 油电混合车辆路径规划、 船舶航迹规划、 全路径规划规划、 仓储巡逻、公交车时间调度、水库调度优化、多式联运优化等等~小众优化类生产调度、经济调度、装配线调度、充电优化、车间调度、发车优化、水库调度、三维装箱、物流选址、货位优化、公交排班优化、充电桩布局优化、车间布局优化、集装箱船配载优化、水泵组合优化、解医疗资源分配优化、设施布局优化、可视域基站和无人机选址优化、背包问题、 风电场布局、时隙分配优化、 最佳分布式发电单元分配、多阶段管道维修、 工厂-中心-需求点三级选址问题、 应急生活物质配送中心选址、 基站选址、 道路灯柱布置、 枢纽节点部署、 输电线路台风监测装置、 集装箱调度、 机组优化、 投资优化组合、云服务器组合优化、 天线线性阵列分布优化、CVRP问题、VRPPD问题、多中心VRP问题、多层网络的VRP问题、多中心多车型的VRP问题、 动态VRP问题、双层车辆路径规划2E-VRP、充电车辆路径规划EVRP、油电混合车辆路径规划、混合流水车间问题、 订单拆分调度问题、 公交车的调度排班优化问题、航班摆渡车辆调度问题、选址路径规划问题、港口调度、港口岸桥调度、停机位分配、机场航班调度、泄漏源定位、冷链、时间窗、多车场等、选址优化、港口岸桥调度优化、交通阻抗、重分配、停机位分配、机场航班调度、通信上传下载分配优化、微电网优化、无功优化、配电网重构、储能配置、有序充电、MPPT优化、家庭用电、电/冷/热负荷预测、电力设备故障诊断、电池管理系统BMSSOC/SOH估算粒子滤波/卡尔曼滤波、 多目标优化在电力系统调度中的应用、光伏MPPT控制算法改进扰动观察法/电导增量法、电动汽车充放电优化、微电网日前日内优化、储能优化、家庭用电优化、供应链优化\智能电网分布式能源经济优化调度虚拟电厂能源消纳风光出力控制策略多目标优化博弈能源调度鲁棒优化等等均可~ 无人机应用方面无人机路径规划、无人机控制、无人机编队、无人机协同、无人机任务分配、无人机安全通信轨迹在线优化、车辆协同无人机路径规划通信方面传感器部署优化、通信协议优化、路由优化、目标定位优化、Dv-Hop定位优化、Leach协议优化、WSN覆盖优化、组播优化、RSSI定位优化、水声通信、通信上传下载分配信号处理方面信号识别、信号加密、信号去噪、信号增强、雷达信号处理、信号水印嵌入提取、肌电信号、脑电信号、信号配时优化、心电信号、DOA估计、编码译码、变分模态分解、管道泄漏、滤波器、数字信号处理传输分析去噪、数字信号调制、误码率、信号估计、DTMF、信号检测电力系统方面 微电网优化、无功优化、配电网重构、储能配置、有序充电、MPPT优化、家庭用电、电/冷/热负荷预测、电力设备故障诊断、电池管理系统BMSSOC/SOH估算粒子滤波/卡尔曼滤波、 多目标优化在电力系统调度中的应用、光伏MPPT控制算法改进扰动观察法/电导增量法、电动汽车充放电优化、微电网日前日内优化、储能优化、家庭用电优化、供应链优化\智能电网分布式能源经济优化调度虚拟电厂能源消纳风光出力控制策略多目标优化博弈能源调度鲁棒优化原创改进优化算法适合需要创新的同学原创改进2025年的波动光学优化算法WOO以及三国优化算法TKOA、白鲸优化算法BWO等任意优化算法均可保证测试函数效果一般可直接核心