283 lines
8.5 KiB
Matlab
283 lines
8.5 KiB
Matlab
%% 功能：单人呼吸心跳原始数据采集与MATLAB处理
|
||
%% 基于平台 IWR1642EVM+DCA1000
|
||
%% 思考：如何将IWR1843EVM+DCA1000采集得到的数据移植到这份代码上
|
||
%% ========================================================================
|
||
clc;clear;close all;
|
||
%% =========================================================================
|
||
%% 读取数据部分
|
||
numADCSamples = 200; % number of ADC samples per chirp
|
||
numADCBits = 16; % number of ADC bits per sample
|
||
numRX = 4; % number of receivers
|
||
numLanes = 2; % do not change. number of lanes is always 2
|
||
isReal = 0; % set to 1 if real only data, 0 if complex data0
|
||
chirpLoop = 2;
|
||
|
||
%% 雷达参数设置
|
||
Fs=4e6; %ADC采样率 见配置说明
|
||
c=3*1e8; %光速
|
||
ts=numADCSamples/Fs;%ADC采样时间
|
||
slope=70e12; %调频斜率
|
||
B_valid =ts*slope; %有效带宽
|
||
detaR=c/(2*B_valid);%距离分辨率
|
||
|
||
%% 读取Bin文件
|
||
Filename = 'xxx.bin'; %文件名
|
||
fid = fopen(Filename,'r');
|
||
adcDataRow = fread(fid, 'int16');
|
||
if numADCBits ~= 16
|
||
l_max = 2^(numADCBits-1)-1;
|
||
adcDataRow(adcDataRow > l_max) = adcDataRow(adcDataRow > l_max) - 2^numADCBits;
|
||
end
|
||
fclose(fid);
|
||
|
||
fileSize = size(adcDataRow, 1);
|
||
PRTnum = fix(fileSize/(numADCSamples*numRX));
|
||
fileSize = PRTnum * numADCSamples*numRX;
|
||
adcData = adcDataRow(1:fileSize);
|
||
% real data reshape, filesize = numADCSamples*numChirps
|
||
if isReal
|
||
numChirps = fileSize/numADCSamples/numRX;
|
||
LVDS = zeros(1, fileSize);
|
||
%create column for each chirp
|
||
LVDS = reshape(adcData, numADCSamples*numRX, numChirps);
|
||
%each row is data from one chirp
|
||
LVDS = LVDS.';
|
||
else
|
||
numChirps = fileSize/2/numADCSamples/numRX; %含有实部虚部除以2
|
||
LVDS = zeros(1, fileSize/2);
|
||
%combine real and imaginary part into complex data
|
||
%read in file: 2I is followed by 2Q
|
||
counter = 1;
|
||
for i=1:4:fileSize-1
|
||
LVDS(1,counter) = adcData(i) + sqrt(-1)*adcData(i+2);
|
||
LVDS(1,counter+1) = adcData(i+1)+sqrt(-1)*adcData(i+3); counter = counter + 2;
|
||
end
|
||
% create column for each chirp
|
||
LVDS = reshape(LVDS, numADCSamples*numRX, numChirps);
|
||
%each row is data from one chirp
|
||
LVDS = LVDS.';
|
||
end
|
||
|
||
%% 重组数据
|
||
adcData = zeros(numRX,numChirps*numADCSamples);
|
||
for row = 1:numRX
|
||
for i = 1:numChirps
|
||
adcData(row, (i-1)*numADCSamples+1:i*numADCSamples) = LVDS(i, (row-1)*numADCSamples+1:row*numADCSamples);
|
||
end
|
||
end
|
||
|
||
retVal= reshape(adcData(1, :), numADCSamples, numChirps); %取第二个接收天线数据，数据存储方式为一个chirp一列
|
||
|
||
process_adc = retVal(:,1:4:end);
|
||
|
||
|
||
%% 距离维FFT（1个chirp)
|
||
% figure;
|
||
% plot((1:numADCSamples)*detaR,db(abs(fft(process_adc(:,1)))));
|
||
% xlabel('距离（m）');
|
||
% ylabel('幅度(dB)');
|
||
% title('距离维FFT（1个chirp）');
|
||
% figure;
|
||
% plot(db(abs(fft(process_adc(:,1)))))
|
||
|
||
%% 相位解缠绕部分
|
||
RangFFT = 512;
|
||
fft_data_last = zeros(1,RangFFT);
|
||
range_max = 0;
|
||
adcdata = process_adc;
|
||
numChirps = size(adcdata, 2);
|
||
|
||
%% 距离维FFT
|
||
fft_data = fft(adcdata,RangFFT);
|
||
fft_data = fft_data.';
|
||
fft_data_abs = abs(fft_data);
|
||
fft_data_abs(:,1:4)=0; %去除直流分量
|
||
real_data = real(fft_data);
|
||
imag_data = imag(fft_data);
|
||
|
||
|
||
for i = 1:numChirps
|
||
for j = 1:RangFFT %对每一个距离点取相位 extract phase
|
||
angle_fft(i,j) = atan2(imag_data(i, j),real_data(i, j));
|
||
end
|
||
end
|
||
|
||
% Range-bin tracking 找出能量最大的点，即人体的位置
|
||
for j = 1:RangFFT
|
||
for i = 1:numChirps % 进行非相干积累
|
||
fft_data_last(j) = fft_data_last(j) + fft_data_abs(i,j);
|
||
end
|
||
if ( fft_data_last(j) > range_max)
|
||
range_max = fft_data_last(j);
|
||
max_num = j;
|
||
end
|
||
end
|
||
|
||
%% 取出能量最大点的相位 extract phase from selected range bin
|
||
angle_fft_last = angle_fft(:,max_num);
|
||
|
||
%% 进行相位解缠 phase unwrapping(手动解)，自动解可以采用MATLAB自带的函数unwrap()
|
||
n = 1;
|
||
for i = 1+1:numChirps
|
||
diff = angle_fft_last(i) - angle_fft_last(i-1);
|
||
if diff > pi
|
||
angle_fft_last(i:end) = angle_fft_last(i:end) - 2*pi;
|
||
n = n + 1;
|
||
elseif diff < -pi
|
||
angle_fft_last(i:end) = angle_fft_last(i:end) + 2*pi;
|
||
end
|
||
end
|
||
|
||
%% phase difference 相位差分后的数据
|
||
angle_fft_last2=zeros(1,numChirps);
|
||
for i = 1:numChirps-1
|
||
angle_fft_last2(i) = angle_fft_last(i+1) - angle_fft_last(i);
|
||
angle_fft_last2(numChirps)=angle_fft_last(numChirps)-angle_fft_last(numChirps-1);
|
||
end
|
||
|
||
% figure;
|
||
% plot(angle_fft_last2);
|
||
% xlabel('点数（N）');
|
||
% ylabel('相位');
|
||
% title('相位差分后的结果');
|
||
|
||
%% IIR带通滤波 Bandpass Filter 0.1-0.6hz，得到呼吸的数据
|
||
fs =20; %呼吸心跳信号采样率
|
||
COE1=chebyshev_IIR; %采用fdatool生成函数
|
||
save coe1.mat COE1;
|
||
breath_data = filter(COE1,angle_fft_last2);
|
||
|
||
% figure;
|
||
% plot(breath_data);
|
||
% xlabel('时间/点数');
|
||
% ylabel('幅度');
|
||
% title('呼吸时域波形');
|
||
|
||
%% 谱估计 -FFT -Peak interval
|
||
N1=length(breath_data);
|
||
fshift = (-N1/2:N1/2-1)*(fs/N1); % zero-centered frequency
|
||
breath_fre = abs(fftshift(fft(breath_data))); %--FFT
|
||
% figure;
|
||
% plot(fshift,breath_fre);
|
||
% xlabel('频率（f/Hz）');
|
||
% ylabel('幅度');
|
||
% title('呼吸信号FFT ');
|
||
|
||
breath_fre_max = 0; % 呼吸频率
|
||
|
||
breath_index1 = length(breath_fre)/2;
|
||
|
||
for i = 1:breath_index1%谱峰最大值搜索 对称其实可以取一半
|
||
if (breath_fre(i) > breath_fre_max)
|
||
breath_fre_max = breath_fre(i);
|
||
if(breath_fre_max<1e-2) %幅度置信 判断是否是存在人的呼吸
|
||
breath_index=numChirps+1;
|
||
else
|
||
breath_index=i;
|
||
end
|
||
end
|
||
end
|
||
|
||
breath_count =(fs*(breath_index1-breath_index)/numChirps)*60; %呼吸频率解算
|
||
|
||
%% IIR带通滤波 Bandpass Filter 0.8-2hz 得到心跳的数据
|
||
COE2=chebyshev_IIR2;
|
||
save coe2.mat COE2;
|
||
heart_data = filter(COE2,angle_fft_last2);
|
||
|
||
N1=length(heart_data);
|
||
fshift = (-N1/2:N1/2-1)*(fs/N1); % zero-centered frequency
|
||
heart_fre = abs(fftshift(fft(heart_data)));
|
||
% figure;
|
||
% plot(fshift,heart_fre);
|
||
% xlabel('频率（f/Hz）');
|
||
% ylabel('幅度');
|
||
% title('心跳信号FFT');
|
||
|
||
heart_fre_max = 0;
|
||
heart_index1 = length(heart_fre)/2 ;
|
||
|
||
for i = 1:heart_index1
|
||
if (heart_fre(i) > heart_fre_max)
|
||
heart_fre_max = heart_fre(i);
|
||
if(heart_fre_max<1e-2)%幅度置信 判断是否是存在人的心跳
|
||
heart_index=numChirps+1;
|
||
else
|
||
heart_index=i;
|
||
end
|
||
end
|
||
end
|
||
heart_count =(fs*(heart_index1-(heart_index-1))/numChirps)*60%心跳频率解算
|
||
|
||
% 2399个帧，约为120s，
|
||
% 如果数据长度够长，则雷达会51.2s对呼吸数据和心跳数据进行一次刷新，
|
||
%以便实现更为精确的检测。
|
||
|
||
disp(['呼吸：',num2str(breath_count),' 心跳：',num2str(heart_count)])
|
||
|
||
%% 动画演示
|
||
%读入波形
|
||
breath_fre;
|
||
heart_fre;
|
||
count =0;
|
||
%显示波形长度
|
||
L =500;
|
||
T_frame =0.05 ;%50ms
|
||
for i = 1:length(breath_fre)-L-1
|
||
figure(10);
|
||
count=count+i;
|
||
[breath_pks,breath_locs] = findpeaks(breath_data(i:i+L )) ;
|
||
breath_fre =(60/ (L*T_frame))*length(breath_pks);
|
||
|
||
[heart_pks,heart_locs] = findpeaks(heart_data(i:i+L )) ;
|
||
heart_fre =(60/ (L*T_frame))*length(heart_pks);
|
||
|
||
if count>500
|
||
subplot(121);
|
||
plot((i:i+L)*T_frame,breath_data(i:i+L ),'g');
|
||
grid on
|
||
xlim([i,i+L]*T_frame);
|
||
ylim([-0.8,0.8]);
|
||
hold on
|
||
xlabel('时间（s）');
|
||
ylabel('幅度');
|
||
title(['呼吸时域波形',num2str(breath_fre),'次']);
|
||
hold off
|
||
|
||
subplot(122);
|
||
plot((i:i+L)*T_frame,heart_data(i:i+L ),'r');
|
||
grid on
|
||
xlim([i,i+L]*T_frame);
|
||
ylim([-0.8,0.8]);
|
||
hold on
|
||
xlabel('时间（s）');
|
||
ylabel('幅度');
|
||
title(['心跳时域波形:',num2str(heart_fre),'次']);
|
||
hold off
|
||
else
|
||
|
||
%呼吸时域
|
||
subplot(121);
|
||
plot((i:i+L)*T_frame,breath_data(i:i+L ),'g');grid on
|
||
xlim([i,i+L]*T_frame);
|
||
ylim([-0.8,0.8]);
|
||
hold on
|
||
xlabel('时间（s）');
|
||
ylabel('幅度');
|
||
title(['呼吸时域波形:',num2str(breath_fre),'次']);
|
||
hold off
|
||
|
||
subplot(122);
|
||
plot((i:i+L)*T_frame,heart_data(i:i+L ),'r');grid on
|
||
xlim([i,i+L]*T_frame);
|
||
ylim([-0.8,0.8]);
|
||
hold on
|
||
xlabel('时间（s）');
|
||
ylabel('幅度');
|
||
title(['心跳时域波形:',num2str(heart_fre),'次']);
|
||
end
|
||
hold off
|
||
end
|
||
|
||
|