DF创客社区发现DF创客商城教程产品资料库

无源管道流体脉动采集装置（ESP32‑C5 + SEN0209）

2026-10-102
Lumen微光
Lumen微光

UID: 19993

11文章
2粉丝
22获取喜欢

管道内部流体流动会产生压力脉动，脉动信号包含流量变化、湍流、振动、泵体异常等特征。传统管道流体监测设备体积大、成本高，本项目使用ESP32‑C5搭配SEN0209振动传感器，以800Hz固定采样频率采集管道壁振动信号，通过滑动平均滤波与快速傅里叶变换（FFT）把时域振动波形转换成频谱，再通过内置Web网页实时展示波形，支持最多8组历史工况本地保存对比，实现低成本无源式管道流体脉动在线采集与特征分析。

 

-“无源”含义：本装置只贴附在管道外壁，不需要开孔、不需要侵入管道内部，依靠拾取管道壁传导的流体振动信号完成监测，属于非侵入式采集。

 

创作环境

 

硬件

ESP32‑C5开发板、SEN0209压电振动传感器、USB 5V供电

接线：SEN0209 → VCC接3.3V，GND接GND，AO模拟输出接入ESP32‑C5 GPIO4，传感器紧贴管道外壁，实现非侵入式采集管壁传导的流体振动信号。

无源管道流体脉动采集装置（ESP32‑C5 + SEN0209）_image_1.webp

选择本产品的原因：

柔性PVDF压电薄膜，贴合曲面管道更友好；带数字/模拟双输出、灵敏度调节电位器，内置电荷放大模块，频率响应范围

 

软件

Arduino IDE 2.x，esp32板包≥3.0.0，依赖库：arduinoFFT；通信方式：ESP32内置WebServer，浏览器Chart.js绘图，LocalStorage本地存储最多8组工况。

 

核心参数：采样频率800Hz，单帧256点，频率分辨率3.125Hz，5点滑动平均，汉明窗FFT。

 

主要原理介绍

 

管道内部流体产生的压力脉动会传递到管道外壁，SEN0209压电传感器将管壁机械振动转换成模拟电压信号；ESP32‑C5使用MCPWM硬件定时器产生稳定800Hz定时中断，采集ADC信号，经过滑动平均降噪，再用FFT将时域波形转为频谱，提取流体脉动的主要特征频率。
网页端可以实时看到时域波形与频谱，支持多工况频谱叠加对比，同时增加硬件启停控制，网页点暂停时，ESP32会关闭MCPWM定时器，停止采样，降低主控负载。

 

代码编写

 

第1步：引入头文件、定义基础配置

 

首先导入WiFi、WebServer、FFT库以及MCPWM驱动，定义硬件引脚、采样频率、FFT点数，预留全局数组用来存放原始采样、滤波后、FFT虚部实部数据。

 

#include <WiFi.h>
#include <WebServer.h>
#include <arduinoFFT.h>
#include "driver/mcpwm.h"
#include "freertos/portmacro.h"

// ========= 用户配置区 =========
const char* WIFI_SSID = "你的WiFi名称";
const char* WIFI_PASS = "你的WiFi密码";
#define SEN0209_PIN     4
#define SAMPLING_FREQ   800
#define FFT_N           256
#define FILTER_SIZE     5
// ==============================

WebServer server(80);
arduinoFFT FFT = arduinoFFT();

double vReal[FFT_N];
double vImag[FFT_N];
double rawBuf[FFT_N];
double filterBuf[FFT_N];
float ADC_MID = 2048.0f;
uint32_t frameId = 0;

volatile bool adcSampleRequest = false;
volatile bool frameReady = false;
volatile bool processingReady = false;
int sampleIndex = 0;
bool samplingEnabled = false;

 

要点： volatile  标记中断中会修改的变量，防止编译器优化掉； samplingEnabled  就是后面用来控制硬件启停的总开关。

 

第2步：MCPWM定时中断 + 采样启停函数

 

MCPWM产生稳定800Hz定时，中断里面只置标记，不在中断里面读取ADC，避免中断里面执行耗时操作；新增startSampling / stopSampling，用来启动、关闭MCPWM硬件定时器，同时复位采样状态。

 

void IRAM_ATTR mcpwmIsrHandler(void *arg)
{
  adcSampleRequest = true;
  portYIELD_FROM_ISR(pdFALSE);
}

void initMcpwmTimer(uint32_t sampleFreq)
{
  mcpwm_config_t mcpwmCfg = {
    .frequency = sampleFreq,
    .duty_cycle = 50.0f,
    .counter_mode = MCPWM_UP_COUNTER,
    .carrier_mode = MCPWM_CARRIER_DISABLE
  };
  mcpwm_init(MCPWM_UNIT_0, MCPWM_TIMER_0, &mcpwmCfg);
  mcpwm_isr_register(MCPWM_UNIT_0, MCPWM_TIMER_0, MCPWM_OPR_A, mcpwmIsrHandler, NULL);
  Serial.printf("[MCPWM] Timer config done, sampling freq: %d Hz\n", sampleFreq);
}

void startSampling()
{
  if(!samplingEnabled)
  {
    mcpwm_start(MCPWM_UNIT_0, MCPWM_TIMER_0);
    samplingEnabled = true;
    Serial.println("[Sampling] START");
  }
}

void stopSampling()
{
  if(samplingEnabled)
  {
    mcpwm_stop(MCPWM_UNIT_0, MCPWM_TIMER_0);
    samplingEnabled = false;
    adcSampleRequest = false;
    frameReady = false;
    processingReady = false;
    sampleIndex = 0;
    Serial.println("[Sampling] STOP");
  }
}

 

第3步：信号预处理函数（滑动平均滤波 + ADC零点标定）

 

SEN0209静态时输出一个固定基线电压，ADC零点标定用来自动测出基线，后续采样减去基线，把静态值归零；5点滑动平均用来抑制ADC高频随机噪声。

 

double slidingFilter(double *src, int idx)
{
  double sum = 0;
  int cnt = 0;
  for(int k = idx - ((FILTER_SIZE-1)/2); k <= idx + ((FILTER_SIZE-1)/2); k++)
  {
    if(k >= 0 && k < FFT_N)
    {
      sum += src[k];
      cnt++;
    }
  }
  return sum / cnt;
}

void calibrateADC()
{
  Serial.println("[Calibrate] Start ADC zero calibration, keep SEN0209 static...");
  float sum = 0.0f;
  for(int i = 0; i < 30; i++)
  {
    sum += analogRead(SEN0209_PIN);
    delay(10);
  }
  ADC_MID = sum / 30.0f;
  Serial.printf("[Calibrate] ADC mid value = %.2f\n", ADC_MID);
}

 

第4步：ADC采集与FFT运算

 

hardwareSampleTask  在主循环执行，当MCPWM中断标记到来时读取ADC，填满256点一整帧后标记  frameReady=true ； processFFTFrame  执行滑动平均、汉明窗、FFT正向计算、频谱幅值归一化，完成一帧频域转换。

 

void hardwareSampleTask()
{
  if(!samplingEnabled) return;
  if(adcSampleRequest)
  {
    adcSampleRequest = false;
    if(sampleIndex < FFT_N)
    {
      rawBuf[sampleIndex] = analogRead(SEN0209_PIN) - ADC_MID;
      sampleIndex++;
      if(sampleIndex >= FFT_N)
      {
        frameReady = true;
        sampleIndex = 0;
      }
    }
  }
}

void processFFTFrame()
{
  for(int i = 0; i < FFT_N; i++)
  {
    filterBuf[i] = slidingFilter(rawBuf, i);
    vReal[i] = filterBuf[i];
    vImag[i] = 0;
  }
  FFT.Windowing(vReal, FFT_N, FFT_WIN_TYP_HAMMING, FFT_FORWARD);
  FFT.Compute(vReal, vImag, FFT_N, FFT_FORWARD);
  FFT.ComplexToMagnitude(vReal, vImag, FFT_N);
  for(int i = 0; i < FFT_N/2; i++)
  {
    vReal[i] = vReal[i] / (FFT_N / 2.0);
  }
  frameId++;
  processingReady = true;
}

 

第5步：辅助函数，提取主频、拼接JSON

 

getTop3FreqString  遍历频谱数组，找到幅值最大前3个频率作为管道脉动特征主频； getTimeJSON  /  getFftJSON  将时域、频谱数组转为JSON字符串，用于网页传输。

 

String getTop3FreqString()
{
  double freqRes = (double)SAMPLING_FREQ / FFT_N;
  static double val[FFT_N/2];
  static int idxList[3];
  for(int i = 0; i < FFT_N/2; i++) val[i] = vReal[i];
  for(int n = 0; n < 3; n++)
  {
    double maxV = -1;
    int pos = 0;
    for(int i = 0; i < FFT_N/2; i++)
    {
      if(val[i] > maxV)
      {
        maxV = val[i];
        pos = i;
      }
    }
    idxList[n] = pos;
    val[pos] = -99999;
  }
  char buf[128];
  snprintf(buf, sizeof(buf), "%.2f Hz, %.2f Hz, %.2f Hz",
    idxList[0] * freqRes,
    idxList[1] * freqRes,
    idxList[2] * freqRes
  );
  return String(buf);
}

String getTimeJSON()
{
  String s = "[";
  for(int i = 0; i < FFT_N; i++)
  {
    s += String(filterBuf[i]);
    if(i != FFT_N -1) s += ",";
  }
  s += "]";
  return s;
}

String getFftJSON()
{
  String s = "[";
  for(int i = 0; i < FFT_N/2; i++)
  {
    s += String(vReal[i]);
    if(i != (FFT_N/2)-1) s += ",";
  }
  s += "]";
  return s;
}

 

第6步：Web路由、内嵌前端网页

 

定义Web服务路由：

  •  /  返回完整前端网页，包含Chart.js绘图、工况保存、启停按钮
  • ​/data  返回最新一帧时域、频谱、前3主频JSON数据
  • ​/start 、 /stop  接收前端指令，调用上面  startSampling()  /  stopSampling() ，实现硬件启停同步

 

void handleRoot()
{
  String html = R"HTML(
<!DOCTYPE html>
<html lang="zh-CN">
<head>
<meta charset="UTF-8">
<title>管道流体脉动采集 SEN0209</title>
<script src="https://cdn.jsdelivr.net/npm/chart.js"></script>
<style>
body{font-family:system-ui;margin:12px;}
.btn{padding:8px 14px;margin:4px;cursor:pointer}
.line{margin:8px 0;}
</style>
</head>
<body>
<h2>无源管道流体脉动采集装置 ESP32-C5 + SEN0209</h2>
<div class="line"><button class="btn" id="btnToggle" onclick="toggleCollect()">开始采集</button></div>
<div class="line">
  <button class="btn" onclick="saveCurrent()">保存当前工况</button>
  <button class="btn" onclick="clearSaved()">清空已保存工况</button>
</div>
<div class="line">
  <button class="btn" onclick="clearAllCanvas()">清空全部界面曲线</button>
  <button class="btn" onclick="clearUnsaved()">清空未保存曲线</button>
</div>
<div id="topFreq">特征频率：等待采集</div>
<div style="height:320px"><canvas id="timeChart"></canvas></div>
<div style="height:320px"><canvas id="fftChart"></canvas></div>
<script>
const timeCtx = document.getElementById('timeChart').getContext('2d');
const fftCtx = document.getElementById('fftChart').getContext('2d');
const freqRes = 3.1250;
const sampleDt = 0.001250;
const colorList = ['#16a34a','#f59e0b','#8b5cf6','#06b6d4','#f43f5e','#64748b','#84cc16','#ec4899'];
const MAX_CASE = 8;
let lastFrameId = -1;
let timeChart = new Chart(timeCtx, {type:'line',data:{labels:[],datasets:[{label:'振动时域',data:[],borderColor:'#2563eb',fill:false,tension:0.1}]},options:{responsive:true,maintainAspectRatio:false,scales:{x:{title:{display:true,text:'时间 s'}},y:{title:{display:true,text:'ADC偏移值'}}}}});
let fftChart = new Chart(fftCtx, {type:'line',data:{labels:[],datasets:[{label:'实时频谱',data:[],borderColor:'#db2777',fill:false,tension:0.1}]},options:{responsive:true,maintainAspectRatio:false,scales:{x:{title:{display:true,text:'频率 Hz'}},y:{title:{display:true,text:'归一化幅值'}}}}});
function loadSavedCases(){const raw = localStorage.getItem("vibrationSavedCases");if(raw){try{return JSON.parse(raw);}catch(e){localStorage.removeItem("vibrationSavedCases");alert("⚠️ 历史工况损坏，已清空");}}return [];}
function saveCasesToLocalStorage(arr){localStorage.setItem("vibrationSavedCases", JSON.stringify(arr));}
let savedDatasets = loadSavedCases();
for(let k=0; k < savedDatasets.length; k++){fftChart.data.datasets.push({label:'工况'+(k+1),data:savedDatasets[k],borderColor:colorList[k],fill:false,tension:0.1});}fftChart.update();
let isFetching = false;let isRunning = false;let timerId = null;
async function fetchData(){const r = await fetch("/data");const j = await r.json();return j;}
async function espStart(){await fetch("/start");}
async function espStop(){await fetch("/stop");}
async function collectLoop(){if(!isRunning) return;if(isFetching) return;isFetching = true;try{const d = await fetchData();if(d.frameId === lastFrameId){isFetching = false;timerId = setTimeout(collectLoop,350);return;}lastFrameId = d.frameId;const tLabels = d.time.map((v,i)=> (i*sampleDt).toFixed(4));timeChart.data.labels = tLabels;timeChart.data.datasets[0].data = d.time;timeChart.update('none');const fLabels = d.fft.map((v,i)=>(i*freqRes).toFixed(2));fftChart.data.labels = fLabels;fftChart.data.datasets[0].data = d.fft;for(let k=0;k<savedDatasets.length;k++){if(fftChart.data.datasets[k+1]){fftChart.data.datasets[k+1].data = savedDatasets[k];}}fftChart.update('none');document.getElementById('topFreq').innerText = "特征频率前3："+d.top3;}catch(e){console.error(e);}isFetching = false;timerId = setTimeout(collectLoop,350);}
async function toggleCollect(){const btn = document.getElementById("btnToggle");if(!isRunning){await espStart();isRunning = true;btn.innerText = "暂停采集";collectLoop();}else{await espStop();isRunning = false;clearTimeout(timerId);btn.innerText = "开始采集";}}
function saveCurrent(){if(savedDatasets.length >= MAX_CASE){alert("最多保存8组工况");return;}const currentFftData = fftChart.data.datasets[0].data;if (!currentFftData || currentFftData.length === 0){alert("实时曲线为空，请先采集");return;}const validData = currentFftData.filter(x=>!isNaN(x));if(validData.length < currentFftData.length * 0.8){alert("频谱有效数据过少，无法保存");return;}const copyData = [...validData];savedDatasets.push(copyData);saveCasesToLocalStorage(savedDatasets);const idx = savedDatasets.length - 1;fftChart.data.datasets.push({label: '工况'+(idx+1),data: copyData,borderColor: colorList[idx],fill:false,tension:0.1});fftChart.update();alert("✅ 已保存工况，共"+savedDatasets.length+"/8");}
function clearSaved(){savedDatasets = [];saveCasesToLocalStorage(savedDatasets);fftChart.data.datasets = [ fftChart.data.datasets[0] ];fftChart.update();alert("已清空浏览器内保存工况");}
function clearAllCanvas(){savedDatasets = [];saveCasesToLocalStorage(savedDatasets);timeChart.data.labels = [];timeChart.data.datasets[0].data = [];timeChart.update();fftChart.data.labels = [];fftChart.data.datasets = [ fftChart.data.datasets[0] ];fftChart.data.datasets[0].data = [];fftChart.update();document.getElementById('topFreq').innerText = "特征频率：等待采集";alert("全部曲线清空");}
function clearUnsaved(){timeChart.data.datasets[0].data = [];timeChart.update();if(fftChart.data.datasets.length > 0){fftChart.data.datasets[0].data = [];}fftChart.update();document.getElementById('topFreq').innerText = "特征频率：等待采集";alert("仅清空实时曲线，历史工况保留");}
</script>
</body>
</html>
)HTML";
  server.send(200, "text/html", html);
}

void handleData()
{
  if(processingReady)
  {
    String json = "{";
    json += "\"frameId\":" + String(frameId) + ",";
    json += "\"time\":" + getTimeJSON() + ",";
    json += "\"fft\":" + getFftJSON() + ",";
    json += "\"top3\":\"" + getTop3FreqString() + "\"";
    json += "}";
    server.send(200, "application/json", json);
    processingReady = false;
  }
  else
  {
    server.send(200, "application/json", "{\"frameId\":0,\"time\":[],\"fft\":[],\"top3\":\"等待计算\"}");
  }
}

void handleStart()
{
  startSampling();
  server.send(200, "text/plain", "ok");
}

void handleStop()
{
  stopSampling();
  server.send(200, "text/plain", "ok");
}

 

第7步：setup初始化 + loop主循环

 

setup里面依次执行串口初始化、ADC配置、WiFi连接、ADC零点标定、MCPWM定时器初始化、注册全部Web路由，启动WebServer；loop循环里面持续处理HTTP请求，执行ADC采集任务，当一帧采集满后执行FFT运算。

 

void setup()
{
  Serial.begin(115200);
  analogSetAttenuation(ADC_11db);
  pinMode(SEN0209_PIN, INPUT);

  Serial.printf("Connecting WiFi: %s\n", WIFI_SSID);
  WiFi.mode(WIFI_STA);
  WiFi.begin(WIFI_SSID, WIFI_PASS);
  while(WiFi.status() != WL_CONNECTED)
  {
    delay(500);
    Serial.print(".");
  }
  Serial.println("\nWiFi connected");
  Serial.print("ESP32-C5 IP: ");
  Serial.println(WiFi.localIP());

  calibrateADC();
  initMcpwmTimer(SAMPLING_FREQ);

  server.on("/", handleRoot);
  server.on("/data", handleData);
  server.on("/start", handleStart);
  server.on("/stop", handleStop);
  server.begin();
  Serial.println("Web server started");
}

void loop()
{
  server.handleClient();
  hardwareSampleTask();
  if(frameReady)
  {
    processFFTFrame();
    frameReady = false;
  }
}

 

完整代码

 

#include <WiFi.h>
#include <WebServer.h>
#include <arduinoFFT.h>
#include "driver/mcpwm.h"
#include "freertos/portmacro.h"

// ========= 用户配置区 =========
const char* WIFI_SSID = "你的WiFi名称";
const char* WIFI_PASS = "你的WiFi密码";
#define SEN0209_PIN     4
#define SAMPLING_FREQ   800
#define FFT_N           256
#define FILTER_SIZE     5
// ==============================

WebServer server(80);
arduinoFFT FFT = arduinoFFT();

double vReal[FFT_N];
double vImag[FFT_N];
double rawBuf[FFT_N];
double filterBuf[FFT_N];
float ADC_MID = 2048.0f;
uint32_t frameId = 0;

volatile bool adcSampleRequest = false;
volatile bool frameReady = false;
volatile bool processingReady = false;
int sampleIndex = 0;
bool samplingEnabled = false;

void IRAM_ATTR mcpwmIsrHandler(void *arg)
{
  adcSampleRequest = true;
  portYIELD_FROM_ISR(pdFALSE);
}

void initMcpwmTimer(uint32_t sampleFreq)
{
  mcpwm_config_t mcpwmCfg = {
    .frequency = sampleFreq,
    .duty_cycle = 50.0f,
    .counter_mode = MCPWM_UP_COUNTER,
    .carrier_mode = MCPWM_CARRIER_DISABLE
  };
  mcpwm_init(MCPWM_UNIT_0, MCPWM_TIMER_0, &mcpwmCfg);
  mcpwm_isr_register(MCPWM_UNIT_0, MCPWM_TIMER_0, MCPWM_OPR_A, mcpwmIsrHandler, NULL);
  Serial.printf("[MCPWM] Timer config done, sampling freq: %d Hz\n", sampleFreq);
}

void startSampling()
{
  if(!samplingEnabled)
  {
    mcpwm_start(MCPWM_UNIT_0, MCPWM_TIMER_0);
    samplingEnabled = true;
    Serial.println("[Sampling] START");
  }
}

void stopSampling()
{
  if(samplingEnabled)
  {
    mcpwm_stop(MCPWM_UNIT_0, MCPWM_TIMER_0);
    samplingEnabled = false;
    adcSampleRequest = false;
    frameReady = false;
    processingReady = false;
    sampleIndex = 0;
    Serial.println("[Sampling] STOP");
  }
}

double slidingFilter(double *src, int idx)
{
  double sum = 0;
  int cnt = 0;
  for(int k = idx - ((FILTER_SIZE-1)/2); k <= idx + ((FILTER_SIZE-1)/2); k++)
  {
    if(k >= 0 && k < FFT_N)
    {
      sum += src[k];
      cnt++;
    }
  }
  return sum / cnt;
}

void calibrateADC()
{
  Serial.println("[Calibrate] Start ADC zero calibration, keep SEN0209 static...");
  float sum = 0.0f;
  for(int i = 0; i < 30; i++)
  {
    sum += analogRead(SEN0209_PIN);
    delay(10);
  }
  ADC_MID = sum / 30.0f;
  Serial.printf("[Calibrate] ADC mid value = %.2f\n", ADC_MID);
}

void hardwareSampleTask()
{
  if(!samplingEnabled) return;
  if(adcSampleRequest)
  {
    adcSampleRequest = false;
    if(sampleIndex < FFT_N)
    {
      rawBuf[sampleIndex] = analogRead(SEN0209_PIN) - ADC_MID;
      sampleIndex++;
      if(sampleIndex >= FFT_N)
      {
        frameReady = true;
        sampleIndex = 0;
      }
    }
  }
}

void processFFTFrame()
{
  for(int i = 0; i < FFT_N; i++)
  {
    filterBuf[i] = slidingFilter(rawBuf, i);
    vReal[i] = filterBuf[i];
    vImag[i] = 0;
  }
  FFT.Windowing(vReal, FFT_N, FFT_WIN_TYP_HAMMING, FFT_FORWARD);
  FFT.Compute(vReal, vImag, FFT_N, FFT_FORWARD);
  FFT.ComplexToMagnitude(vReal, vImag, FFT_N);
  for(int i = 0; i < FFT_N/2; i++)
  {
    vReal[i] = vReal[i] / (FFT_N / 2.0);
  }
  frameId++;
  processingReady = true;
}

String getTop3FreqString()
{
  double freqRes = (double)SAMPLING_FREQ / FFT_N;
  static double val[FFT_N/2];
  static int idxList[3];
  for(int i = 0; i < FFT_N/2; i++) val[i] = vReal[i];
  for(int n = 0; n < 3; n++)
  {
    double maxV = -1;
    int pos = 0;
    for(int i = 0; i < FFT_N/2; i++)
    {
      if(val[i] > maxV)
      {
        maxV = val[i];
        pos = i;
      }
    }
    idxList[n] = pos;
    val[pos] = -99999;
  }
  char buf[128];
  snprintf(buf, sizeof(buf), "%.2f Hz, %.2f Hz, %.2f Hz",
    idxList[0] * freqRes,
    idxList[1] * freqRes,
    idxList[2] * freqRes
  );
  return String(buf);
}

String getTimeJSON()
{
  String s = "[";
  for(int i = 0; i < FFT_N; i++)
  {
    s += String(filterBuf[i]);
    if(i != FFT_N -1) s += ",";
  }
  s += "]";
  return s;
}

String getFftJSON()
{
  String s = "[";
  for(int i = 0; i < FFT_N/2; i++)
  {
    s += String(vReal[i]);
    if(i != (FFT_N/2)-1) s += ",";
  }
  s += "]";
  return s;
}

void handleRoot()
{
  String html = R"HTML(
<!DOCTYPE html>
<html lang="zh-CN">
<head>
<meta charset="UTF-8">
<title>管道流体脉动采集 SEN0209</title>
<script src="https://cdn.jsdelivr.net/npm/chart.js"></script>
<style>
body{font-family:system-ui;margin:12px;}
.btn{padding:8px 14px;margin:4px;cursor:pointer}
.line{margin:8px 0;}
</style>
</head>
<body>
<h2>无源管道流体脉动采集装置 ESP32-C5 + SEN0209</h2>
<div class="line"><button class="btn" id="btnToggle" onclick="toggleCollect()">开始采集</button></div>
<div class="line">
  <button class="btn" onclick="saveCurrent()">保存当前工况</button>
  <button class="btn" onclick="clearSaved()">清空已保存工况</button>
</div>
<div class="line">
  <button class="btn" onclick="clearAllCanvas()">清空全部界面曲线</button>
  <button class="btn" onclick="clearUnsaved()">清空未保存曲线</button>
</div>
<div id="topFreq">特征频率：等待采集</div>
<div style="height:320px"><canvas id="timeChart"></canvas></div>
<div style="height:320px"><canvas id="fftChart"></canvas></div>
<script>
const timeCtx = document.getElementById('timeChart').getContext('2d');
const fftCtx = document.getElementById('fftChart').getContext('2d');
const freqRes = 3.1250;
const sampleDt = 0.001250;
const colorList = ['#16a34a','#f59e0b','#8b5cf6','#06b6d4','#f43f5e','#64748b','#84cc16','#ec4899'];
const MAX_CASE = 8;
let lastFrameId = -1;
let timeChart = new Chart(timeCtx, {type:'line',data:{labels:[],datasets:[{label:'振动时域',data:[],borderColor:'#2563eb',fill:false,tension:0.1}]},options:{responsive:true,maintainAspectRatio:false,scales:{x:{title:{display:true,text:'时间 s'}},y:{title:{display:true,text:'ADC偏移值'}}}}});
let fftChart = new Chart(fftCtx, {type:'line',data:{labels:[],datasets:[{label:'实时频谱',data:[],borderColor:'#db2777',fill:false,tension:0.1}]},options:{responsive:true,maintainAspectRatio:false,scales:{x:{title:{display:true,text:'频率 Hz'}},y:{title:{display:true,text:'归一化幅值'}}}}});
function loadSavedCases(){const raw = localStorage.getItem("vibrationSavedCases");if(raw){try{return JSON.parse(raw);}catch(e){localStorage.removeItem("vibrationSavedCases");alert("⚠️ 历史工况损坏，已清空");}}return [];}
function saveCasesToLocalStorage(arr){localStorage.setItem("vibrationSavedCases", JSON.stringify(arr));}
let savedDatasets = loadSavedCases();
for(let k=0; k < savedDatasets.length; k++){fftChart.data.datasets.push({label:'工况'+(k+1),data:savedDatasets[k],borderColor:colorList[k],fill:false,tension:0.1});}fftChart.update();
let isFetching = false;let isRunning = false;let timerId = null;
async function fetchData(){const r = await fetch("/data");const j = await r.json();return j;}
async function espStart(){await fetch("/start");}
async function espStop(){await fetch("/stop");}
async function collectLoop(){if(!isRunning) return;if(isFetching) return;isFetching = true;try{const d = await fetchData();if(d.frameId === lastFrameId){isFetching = false;timerId = setTimeout(collectLoop,350);return;}lastFrameId = d.frameId;const tLabels = d.time.map((v,i)=> (i*sampleDt).toFixed(4));timeChart.data.labels = tLabels;timeChart.data.datasets[0].data = d.time;timeChart.update('none');const fLabels = d.fft.map((v,i)=>(i*freqRes).toFixed(2));fftChart.data.labels = fLabels;fftChart.data.datasets[0].data = d.fft;for(let k=0;k<savedDatasets.length;k++){if(fftChart.data.datasets[k+1]){fftChart.data.datasets[k+1].data = savedDatasets[k];}}fftChart.update('none');document.getElementById('topFreq').innerText = "特征频率前3："+d.top3;}catch(e){console.error(e);}isFetching = false;timerId = setTimeout(collectLoop,350);}
async function toggleCollect(){const btn = document.getElementById("btnToggle");if(!isRunning){await espStart();isRunning = true;btn.innerText = "暂停采集";collectLoop();}else{await espStop();isRunning = false;clearTimeout(timerId);btn.innerText = "开始采集";}}
function saveCurrent(){if(savedDatasets.length >= MAX_CASE){alert("最多保存8组工况");return;}const currentFftData = fftChart.data.datasets[0].data;if (!currentFftData || currentFftData.length === 0){alert("实时曲线为空，请先采集");return;}const validData = currentFftData.filter(x=>!isNaN(x));if(validData.length < currentFftData.length * 0.8){alert("频谱有效数据过少，无法保存");return;}const copyData = [...validData];savedDatasets.push(copyData);saveCasesToLocalStorage(savedDatasets);const idx = savedDatasets.length - 1;fftChart.data.datasets.push({label: '工况'+(idx+1),data: copyData,borderColor: colorList[idx],fill:false,tension:0.1});fftChart.update();alert("✅ 已保存工况，共"+savedDatasets.length+"/8");}
function clearSaved(){savedDatasets = [];saveCasesToLocalStorage(savedDatasets);fftChart.data.datasets = [ fftChart.data.datasets[0] ];fftChart.update();alert("已清空浏览器内保存工况");}
function clearAllCanvas(){savedDatasets = [];saveCasesToLocalStorage(savedDatasets);timeChart.data.labels = [];timeChart.data.datasets[0].data = [];timeChart.update();fftChart.data.labels = [];fftChart.data.datasets = [ fftChart.data.datasets[0] ];fftChart.data.datasets[0].data = [];fftChart.update();document.getElementById('topFreq').innerText = "特征频率：等待采集";alert("全部曲线清空");}
function clearUnsaved(){timeChart.data.datasets[0].data = [];timeChart.update();if(fftChart.data.datasets.length > 0){fftChart.data.datasets[0].data = [];}fftChart.update();document.getElementById('topFreq').innerText = "特征频率：等待采集";alert("仅清空实时曲线，历史工况保留");}
</script>
</body>
</html>
)HTML";
  server.send(200, "text/html", html);
}

void handleData()
{
  if(processingReady)
  {
    String json = "{";
    json += "\"frameId\":" + String(frameId) + ",";
    json += "\"time\":" + getTimeJSON() + ",";
    json += "\"fft\":" + getFftJSON() + ",";
    json += "\"top3\":\"" + getTop3FreqString() + "\"";
    json += "}";
    server.send(200, "application/json", json);
    processingReady = false;
  }
  else
  {
    server.send(200, "application/json", "{\"frameId\":0,\"time\":[],\"fft\":[],\"top3\":\"等待计算\"}");
  }
}

void handleStart()
{
  startSampling();
  server.send(200, "text/plain", "ok");
}

void handleStop()
{
  stopSampling();
  server.send(200, "text/plain", "ok");
}

void setup()
{
  Serial.begin(115200);
  analogSetAttenuation(ADC_11db);
  pinMode(SEN0209_PIN, INPUT);

  Serial.printf("Connecting WiFi: %s\n", WIFI_SSID);
  WiFi.mode(WIFI_STA);
  WiFi.begin(WIFI_SSID, WIFI_PASS);
  while(WiFi.status() != WL_CONNECTED)
  {
    delay(500);
    Serial.print(".");
  }
  Serial.println("\nWiFi connected");
  Serial.print("ESP32-C5 IP: ");
  Serial.println(WiFi.localIP());

  calibrateADC();
  initMcpwmTimer(SAMPLING_FREQ);

  server.on("/", handleRoot);
  server.on("/data", handleData);
  server.on("/start", handleStart);
  server.on("/stop", handleStop);
  server.begin();
  Serial.println("Web server started");
}

void loop()
{
  server.handleClient();
  hardwareSampleTask();
  if(frameReady)
  {
    processFFTFrame();
    frameReady = false;
  }
}

 

用户操作流程

  1. 硬件接线：SEN0209接ESP32‑C5 3.3V、GND、GPIO4，传感器紧贴待测管道外壁
  2. ​修改代码WiFi参数：填入你的2.4G WiFi名称与密码，选择ESP32C5开发板编译上传
  3. ​上电标定：上电保持传感器静止，等待ADC零点标定，串口监视器记录ESP32局域网IP
  4. ​网页访问：手机/电脑连同一局域网，浏览器输入ESP32‑C5 IP打开项目页面
  5. ​开始采集：点击【开始采集】，ESP32启动MCPWM硬件采样，页面展示时域波形、FFT频谱与前3个特征频率
  6. ​工况对比：频谱稳定时点击【保存当前工况】，最多8组历史频谱叠加，对比不同管道工况差异
  7. ​曲线管理：按需选择清空已保存工况 / 清空全部曲线 / 仅清空实时曲线
  8. ​暂停实验：点击【暂停采集】，ESP32关闭MCPWM定时器，停止ADC采集和FFT运算，降低主控负载，再次点击恢复采集

 

展示与技巧

点击开始时开始采集数据

无源管道流体脉动采集装置（ESP32‑C5 + SEN0209）_image_2.webp

可以随时点击暂停查看数据

保存工况后可通过点击图标来隐藏多余曲线

无源管道流体脉动采集装置（ESP32‑C5 + SEN0209）_image_3.webp

无源管道流体脉动采集装置（ESP32‑C5 + SEN0209）_image_4.webp

还可以通过第三行清空按钮来进行数据清理

无源管道流体脉动采集装置（ESP32‑C5 + SEN0209）_image_5.webp

 

装置布置

无源管道流体脉动采集装置（ESP32‑C5 + SEN0209）_image_6.webp

 

本套装置主要适用范围

例：

教学科创实验：课堂/科创比赛，演示流体脉动、管壁振动、FFT频谱分析原理

 

工业现场做初步筛查、对比试验，但不建议直接当作工业计量、安全联锁的正式测量设备。

创作许可协议

本项目采用 CC BY(署名) 进行许可。

教程ESP32造点好玩的

相关推荐

评论(0)
- 没有更多了 -
Lumen微光
Lumen微光

UID: 19993

11文章
2粉丝
22获取喜欢

创作许可协议

本项目采用 CC BY(署名) 进行许可。

相关推荐

手机床电机超时后为什么再按一次也不动
手机床电机超时后为什么再按一次也不动不脱发的程序猿的用户头像不脱发的程序猿30
如何检测和解决I2C通信死锁
如何检测和解决I2C通信死锁不脱发的程序猿的用户头像不脱发的程序猿20
SPI、DSPI、QSPI技术对比
SPI、DSPI、QSPI技术对比不脱发的程序猿的用户头像不脱发的程序猿20