代码拉取完成,页面将自动刷新
classdef LightIntensityEnv < rl.env.MATLABEnvironment
%LIGHTINTENSITYENV: Template for defining custom environment in MATLAB.
%% Properties (set properties' attributes accordingly)
properties
% Specify and initialize environment's necessary properties
% % Action Max:pitch的增量,yaw的增量,
% actionMax = [pi,pi];
%接收平面初始俯仰角为0 度
PitchAngle = 0;
%接收平面初始水平转动的角为 0度
YawAngle= 0;
%接收平面的的法向量量初始时竖直向下
n = [0 0 -1]';
% 接收器位置
rpos = [0 1];
% 两个接收器之间的距离
padding = 0.3;
% 接收器最低光强
IntensityThreshold = 5e-8;
%当前时间
t = 0;
% 采样时间
Ts = 0.02;
% 历史状态集
HisIntensities = [];
HisReward = [];
end
properties
% 初始化系统状态 [intensity1,intensity2]'
State = zeros(2,1)
end
properties(Access = protected)
% Initialize internal flag to indicate episode termination
IsDone = false
end
%% Necessary Methods
methods
function this = LightIntensityEnv()
% Initialize Observation settings.
numObs = 4;
ObservationInfo = rlNumericSpec([numObs 1]);
ObservationInfo.Name = 'observation';
ObservationInfo.Description = 'intensity1,intensity2 pitch yaw';
% Initialize Action settings
numAct = 2;
ActionInfo = rlNumericSpec([numAct 1]);
ActionInfo.Name = 'action';
% 俯仰角和转向角的增量
ActionInfo.Description = 'delta pitch, delta yaw';
% The following line implements built-in functions of RL env
this = this@rl.env.MATLABEnvironment(ObservationInfo,ActionInfo);
rpos = [0.7 0.7];
this.rpos = rpos;
end
% Apply system dynamics and simulates the environment with the
% given action for one step. action is 2x1 matrix
function [Observation,Reward,IsDone,LoggedSignals] = step(this,Action)
LoggedSignals = [];
% Get action
act = saturate(this,Action);
dp = act(1);
dy = act(2);
y = this.YawAngle + dy;
p = this.PitchAngle+dp;
% 旋转后的法向量需要使用如下变换
% 固定⾓坐标系,绕固定轴旋转,角度单位应为度,旋转算⼦乘的顺序从右边开始
R_xyz= rotz(y)*rotx(p);
% 接收器的法向量
n = R_xyz*this.n;
%接收到的光强
intensities = get_intensity(this.rpos,this.t,this.padding,n);
% Update system states
this.PitchAngle = p;
this.YawAngle = y;
%调用本类的函数需要传递this参数。
iLevels = map2level(this,intensities);
Observation = [iLevels(1);iLevels(2);this.PitchAngle;this.YawAngle];
this.State = Observation;
% Check terminal condition
IsDone = any(abs(intensities) < this.IntensityThreshold);
this.IsDone = IsDone;
this.HisIntensities = [this.HisIntensities,this.State];
% Get reward
Reward = getReward(this);
% (optional) use notifyEnvUpdated to signal that the
% environment has been updated (e.g. to update visualization)
notifyEnvUpdated(this);
end
% Reset environment to initial state and output initial observation
function InitialObservation = reset(this)
% "reset env" ,测试每次是否调用reset函数
% 重新初始化
this.YawAngle = 0;
this.PitchAngle = 0;
this.HisIntensities = [];
this.HisReward = [];
% 旋转后的法向量需要使用如下变换
% 固定⾓坐标系,绕固定轴旋转,角度单位应为度,旋转算⼦乘的顺序从右边开始
R_xyz= rotz(this.YawAngle)*rotx(this.PitchAngle);
% 接收器的法向量
n = R_xyz*this.n;
intensities = get_intensity(this.rpos,this.t,this.padding,n);
iLevels = map2level(this,intensities);
InitialObservation = [iLevels(1);iLevels(2);this.PitchAngle;this.YawAngle];
this.State = InitialObservation;
% 角度为0时 1.0e-07 *
% 0.9216
% 0.7314
% (optional) use notifyEnvUpdated to signal that the
% environment has been updated (e.g. to update visualization)
this.HisIntensities = [this.HisIntensities,this.State];
notifyEnvUpdated(this);
end
end
%% Optional Methods (set methods' attributes accordingly)
methods
% Helper methods to create the environment
% Discrete force 1 or 2
function saction = saturate(this,action)
% 没有限制
% if abs(action)>actionMax
% action = sign(axtion)*actionMax;
% end
saction = double(action);
end
function intLevel = map2level(this,intensity)
intLevel = sign(intensity).*log(abs(intensity)/this.IntensityThreshold)*10;
end
% Reward function
function Reward = getReward(this)
if this.IsDone
Reward = -1;
return
end
curstate = this.HisIntensities(:,end);
curstate = curstate(1);
s = size(this);
len = s(2);
Tc = 1;
if len <= Tc
prestate = this.HisIntensities(:,:);
else
prestate = this.HisIntensities(:,end-Tc:end-1);
end
prestate = mean(prestate,2);
prestate = prestate(1);
dIntensity = curstate - prestate;
ir = 10* dIntensity;
sr = curstate(1)/10;
Reward = 10*(ir+sr);
this.HisReward = [this.HisReward;ir,sr];
end
% (optional) Visualization method
function plot(this)
% Initiate the visualization
% Update the visualization
envUpdatedCallback(this)
end
% (optional) Properties validation through set methods
function set.State(this,state)
validateattributes(state,{'numeric'},{'finite','real','vector','numel',4},'','State');
this.State = state(:);
notifyEnvUpdated(this);
end
function set.PitchAngle(this,val)
validateattributes(val,{'numeric'},{'finite','real','scalar'},'','PitchAngle');
this.PitchAngle = val;
notifyEnvUpdated(this);
end
function set.YawAngle(this,val)
validateattributes(val,{'numeric'},{'finite','real','scalar'},'','YawAngle');
this.YawAngle = val;
notifyEnvUpdated(this);
end
end
methods (Access = protected)
% (optional) update visualization everytime the environment is updated
% (notifyEnvUpdated is called)
function envUpdatedCallback(this)
end
end
end
此处可能存在不合适展示的内容,页面不予展示。您可通过相关编辑功能自查并修改。
如您确认内容无涉及 不当用语 / 纯广告导流 / 暴力 / 低俗色情 / 侵权 / 盗版 / 虚假 / 无价值内容或违法国家有关法律法规的内容,可点击提交进行申诉,我们将尽快为您处理。