You are tasked with solving a variation of the Vehicle Routing Problem (VRP) with Time Windows for a delivery fleet. Due to the problem size (1000+ nodes), exact algorithms are infeasible. Design a hybrid metaheuristic algorithm (e.g., combining Simulated Annealing with Genetic Algorithms or Tabu Search). Provide a detailed explanation of: 1) The encoding scheme for the solution (chromosome representation). 2) The fitness function used to evaluate solutions, penalizing violations of time windows and capacity constraints. 3) The crossover and mutation operators specifically tailored for routing problems to maintain valid routes. 4) The convergence criteria and diversification strategies to avoid local optima. Include pseudocode for the main loop.