1

我正在创建一个 android phonegap 应用程序，它使用加速度计来检测道路上的缺陷。但是，当我使用原始加速度计值（x、y 和 z）时，它们从一次尝试到另一次尝试变化太大。我尝试使用相同的设备、相同的车辆和位置来获取相同的缺陷，但是加速度计给出的值完全不同。我每 0.25 秒获取一次加速度计值，所以每秒我会得到 4 个 x、y 和 z 值。

谁能告诉我如何平滑这些值以获得更平滑和更可靠的值？我将需要使用这些值对它们进行进一步处理，因此 x、y 和 z 值的准确性至关重要。是否有任何公式可以应用于这些值以获得更平滑的输出？

更新：添加我的代码片段：

function acc() {
var accOpt = { frequency: 1000 };  
previousReading = {
    x   :   null,
    y   :   null,
    z   :   null,
    lr  :   null,
    fb  :   null            
  }; 
watchID = navigator.accelerometer.watchAcceleration(win, fail, accOpt);    
}

function win (acceleration) {

var tiltLR = 0;
var tiltFB = 0;

var facingUp = -1;
if (acceleration.z > 0) {
  facingUp = +1;
}

tiltLR = Math.round(((acceleration.x) / 9.81) * -90);
tiltFB = Math.round(((acceleration.y + 9.81) / 9.81) * 90 * facingUp);



  var changes = {};

  if (previousReading.x !== null) {
      changes.x = Math.abs(previousReading.x-acceleration.x);
      changes.y = Math.abs(previousReading.y-acceleration.y);
      changes.z = Math.abs(previousReading.z-acceleration.z);
      changes.lr = Math.abs(previousReading.lr-tiltLR);
      changes.fb = Math.abs(previousReading.fb-tiltFB);
  }

  previousReading = {
      x: acceleration.x,
      y: acceleration.y,
      z: acceleration.z,
      lr: tiltLR,
      fb: tiltFB
  };


console.log('previousReading x: ' + previousReading.x + ' previousReading y: ' + previousReading.y + ' previousReading z: ' + previousReading.z+ ' previousReading lr: ' + tiltLR + ' previousReading fb: ' + tiltFB);

console.log('changes x: ' + changes.x  + ' changes y: ' + changes.y + ' changes z: ' + changes.z + ' changes lr: ' + changes.lr + ' changes fb: ' + changes.fb);

console.log(acceleration.x + ' ' + acceleration.y + ' ' + acceleration.z);

 }

过滤器：我遇到过这段代码来平滑 x、y 和 z 值。如果我将此应用于当前的 x、y、z 值，然后将它们与以前的值（也被过滤掉）进行比较，这对我来说是否有效？

 #define kFilteringFactor 0.1

 // Use a basic low-pass filter to keep only the gravity component of each axis.
 grav_accelX = (acceleration.x * kFilteringFactor) + ( grav_accelX * (1.0 - kFilteringFactor));
 grav_accelY = (acceleration.y * kFilteringFactor) + ( grav_accelY * (1.0 - kFilteringFactor));
 grav_accelZ = (acceleration.z * kFilteringFactor) + ( grav_accelZ * (1.0 - kFilteringFactor));

 // Subtract the low-pass value from the current value to get a simplified high-pass filter
 instant_accelX = acceleration.x - ( (acceleration.x * kFilteringFactor) + (instant_accelX * (1.0 - kFilteringFactor)) );
 instant_accelY = acceleration.y - ( (acceleration.y * kFilteringFactor) + (instant_accelY * (1.0 - kFilteringFactor)) );
 instant_accelZ = acceleration.z - ( (acceleration.z * kFilteringFactor) + (instant_accelZ * (1.0 - kFilteringFactor)) );
4

0 回答 0