Algorithm 2: Solving the energy-efficient collision-free machine/AGV scheduling problem in the physical optimization layer
1:	if assigned then task state = optimization
2:	if machine/AGV state = idle
3:	input the best solution in the virtual planning layer
4:	initialize the parameters of energy-efficient collision-free machine/AGV scheduling problem
5:	initialize the parameters of APC
6:	initialize the parameters of the solving system
7:	encode the APC individual by Equation (39)
8:	select fitness function by Equations (37) and (28)
9:	for ite = 1 to ${Ite}_{max}$
10:		produce the random seeds by $p_{seed}$
11:		produce seeds from the previous fruits
12:		calculate fitness by Equations (37) and (28)
13:		choose the best solutions by $p_{grow}$
14:		choose the elite individual with the highest fitness
15:		produce a fruit through parthenogenesis
16:		produce fruits through social learning by $p_{fruit}$
17:		end justification by $e_{th}$
18:	end for
19:	output the best solution in the physical optimization layer
20:	if scheduled then task state = performing
21:	if assigned then machine/AGV state = loading
22:	if scheduled then machine/AGV state = processing/transportation
23:	if detected then machine/AGV state = obstacle avoidance
24:	if detected then machine/AGV state = obstacle avoidance
25:	if avoided then machine/AGV state = processing/transportation
26:	if completed then machine/AGV state = idle