Добавлены изыскания про PID

43 PID control sample/PIDControlR_Initialize.R

@@ -0,0 +1,20 @@
+# Initialization of the PID control simulation 
+
+# controller parameters
+Kp = 10                 # proportional gain
+Ti = 1                  # integral time
+Td = 0.01               # derivative time
+
+# simulation parameters
+dt = .1                 # time step
+tt = seq(0, 100, by=dt) # time vector
+
+# initialize the following to a vector of zeros
+# as long as the time variable tt
+# - PV, process variable
+# - U, control output
+# - E, error
+# - EI, error integral
+# - ED, error derivative
+PV = U = E = EI = ED = rep(0, length(tt))
+PV[1] = 5 # initial state of the process variable

43 PID control sample/PIDControlR_ProcessVariable.R

@@ -0,0 +1,10 @@
+# Implementation of a PID control algorithm in R. A simple process for testing the controller. 
+
+pv = function(pv.prev, u, tt) {
+  out = pv.prev*1.1 + .5           # exponential growth + linear growth 
+  out = out - 0.1*u                # the control response
+  out = out + .5*runif(length(tt)) # a little noise, just for fun
+
+  if (out < 0) out = 0             # keep values positive
+  return(out)
+}

43 PID control sample/PIDControlR_SetpointSinusoid.R

@@ -0,0 +1,2 @@
+# Sinusoidal setpoint profile to test PID control in R 
+SP = sin(0.5*tt) + cos(.8*tt) + 5

43 PID control sample/PIDControlR_SetpointSteps.R

@@ -0,0 +1,4 @@
+# Step change setpoint profile for PID control simulation
+SP = rep(10, length(tt))
+SP[which(tt >= 30)] = 5
+SP[which(tt >= 60)] = 20

43 PID control sample/PIDControlR_SimulationLoop.R

@@ -0,0 +1,13 @@
+# Simulation loop for a PID controller in R f
+
+or (k in 2:length(tt)) {
+  PV[k] = pv(PV[k-1], U[k-1], tt[k])
+  E[k] = PV[k] - SP[k]
+
+  EI[k] = EI[k-1] + E[k]*dt  # integral
+  ED[k] = (E[k] - E[k-1])/dt # derivative
+
+  U[k] = Kp*(E[k] + (1/Ti)*sum(E*dt) + Td*ED[k]) 
+
+  if (U[k] < 0) U[k] = 0
+}