为了教会小学弟PID，我连夜肝出了这篇PID的实操（未完成坑学弟版）

下面废话少说直接步入正题：

首先PID最重要的不过就是调参部分，这里我仍然重复一下我认为所谓的调参口诀吧。

PID中P调大增大响应速度，D调大增加预判，I消除稳差，这三个系数并不是越大越好，P太大会超调,D太大会震荡，i值有累计效应。

这里不懂具体什么是PID的建议看一下我上篇博客自控篇(一)：PID算法的基础了解_靳小锅er的博客-CSDN博客

下面正式开始正题：

下面直接上有关PID代码

void CAM_PID_Controler(void) {
	static int bias_i=0;
	static int last_bias = 0;
	switch(en_line) {
		case front_line: {
			double bias = rou[0] +theta[0] -offset;
			double uk_Motor,uk_Steer;
			bias_i += bias;
			if(bias_i>=3000)bias_i = 1500;
			if(bias_i<=-3000)bias_i = -1500;
			uk_Motor = kp_Motor*bias + ki_Motor*bias_i + kd_Motor*(bias-last_bias);
			uk_Steer = kp_Steer*bias + ki_Steer*bias_i + kd_Steer*(bias-last_bias);
			targetL_A = targetL + uk_Motor;
			targetR_A = targetR - uk_Motor;
			
			pwmSteer2 = targetSteer2 + uk_Steer;
			if(pwmSteer2 > 2000) pwmSteer2 = 1700;
			else if(pwmSteer2 < 1000) pwmSteer2 = 1100;
			last_bias = bias;
			break;
		}
		case back_line: {
			double bias= rou[0] +theta[0] -offset;
			double uk_Motor,uk_Steer;
			bias_i += bias;
			if(bias_i>=3000)bias_i = 1500;
			if(bias_i<=-3000)bias_i = -1500;
			uk_Motor = rp_Motor*bias + /*ri_Motor*bias_i +*/ rd_Motor*(bias-last_bias);
			uk_Steer = rp_Steer*bias + /*ri_Steer*bias_i +*/ rd_Steer*(bias-last_bias);
			targetL_A = targetL + uk_Motor;
			targetR_A = targetR - uk_Motor;
			
			pwmSteer2 = targetSteer2 - uk_Steer;
			if(pwmSteer2 > 2000) pwmSteer2 = 1700;
			else if(pwmSteer2 < 1000) pwmSteer2 = 1100;
			last_bias = bias;
			break;
		}
		case front_left_line: {
			double bias = (rou[0] -fabs(rou[1]))/2 +(theta[0]-fabs(theta[1]))/2 -offset;
			double uk_Motor,uk_Steer;
			bias_i += bias;
			if(bias_i>=3000)bias_i = 1500;
			if(bias_i<=-3000)bias_i = -1500;
			uk_Motor = kp_Motor*bias + ki_Motor*bias_i + kd_Motor*(bias-last_bias);
			uk_Steer = kp_Steer*bias + ki_Steer*bias_i + kd_Steer*(bias-last_bias);
			targetL_A = targetL + uk_Motor;
			targetR_A = targetR - uk_Motor;
			
			pwmSteer2 = targetSteer2 + uk_Steer;
			if(pwmSteer2 > 2000) pwmSteer2 = 1700;
			else if(pwmSteer2 < 1000) pwmSteer2 = 1100;
			last_bias = bias;
			break;
		}
		case front_right_line: {
			double bias = (rou[0] +fabs(rou[1]))/2 +(theta[0]+fabs(theta[1]))/2 -offset;
			double uk_Motor,uk_Steer;
			bias_i += bias;
			if(bias_i>=3000)bias_i = 1500;
			if(bias_i<=-3000)bias_i = -1500;
			uk_Motor = kp_Motor*bias + ki_Motor*bias_i + kd_Motor*(bias-last_bias);
			uk_Steer = kp_Steer*bias + ki_Steer*bias_i + kd_Steer*(bias-last_bias);
			targetL_A = targetL + uk_Motor;
			targetR_A = targetR - uk_Motor;
			
			pwmSteer2 = targetSteer2 + uk_Steer;
			if(pwmSteer2 > 2000) pwmSteer2 = 1700;
			else if(pwmSteer2 < 1000) pwmSteer2 = 1100;
			last_bias = bias;
			break;
		}
		case dis_line:
		default:
		{
			targetL_A = targetL;
			targetR_A = targetR;
			pwmSteer2 = targetSteer2;
			break;
		}
	}
}

 其实其中有关于PID的核心代码很短，不过就是分为三个量，Bias和last_bias还有Bias-last_bias;根据公式kp_Motor*bias + ki_Motor*bias_i + kd_Motor*(bias-last_bias)求出该设置的pwm占空比，这里不再一一赘述。

下面我们开始如何进行PID三个参数的调试。

明天再更

标签：last,double,PID,肝出,pwmSteer2,bias,实操,uk
来源： https://blog.csdn.net/qq_52528413/article/details/121320524