A Study of Network Violator Interception Based on a Reliable Game Model

Algorithm 1

SA on RILP.

Step  1. Initialization.

  (1.1) Randomly select 𝑁 candidate locations to initialize an initial feasible solution 𝑥⁰. For an arbitrary

     route 𝑖, determine ℎ⁰ according to interceptor layout 𝑥⁰, then compute objective function value 𝑧(𝑥⁰)

     according to (6). Let 𝑥=𝑥⁰.

  (1.2) Set Markov length ML, initial temperature 𝑇₀, error 𝜀, and let 𝑇=𝑇₀.

  (1.3) Set the outer iteration counter 𝑛=1.

Step  2. For the given 𝑇, perform the following:

  (2.1) Set the inner iteration counter to 𝑘=1.

  (2.2) Randomly select an element from 𝑥 (selected candidate locations) and 𝜙−𝑥 (unselected candidate

      locations), respectively, and exchange them to produce a new solution 𝑥.

  (2.3) Compute the objective function value 𝑧(𝑥) according to (6).

  (2.4) Set Δ𝑧=𝑧(𝑥)−𝑧(𝑥), if Δ𝑧>0, 𝑥=𝑥; else let 𝑥=𝑥 with probability 𝑝(Δ𝑧)=exp(Δ𝑧/𝑇).

  (2.5) Check whether 𝑘=ML. If 𝑘=ML, go to Step  3; else 𝑘=𝑘+1, return to (2.2).

Step  3. Determine whether to stop

  (3.1) Check whether 𝑇<𝜀. If 𝑇<𝜀, stop; go to (3.2)

  (3.2) Let 𝑇=𝜂𝑇, 𝑛=𝑛+1, and return to Step  2. (where 𝜂 decides the speed of annealing)