Aiming at the problem of the low utilization rate of equipment in the scheduling scheme due to ignoring the processing route constraints in the processing of multi-variety and small-batch complex products,a mathematical model of a multi-process flexible job shop was established with the goal of optimal completion time.Since the complexity of the model increased sharply after adding multiple constraints,an improved atomic orbital search algorithm was proposed to solve the model effectively.In the improved algorithm,a three-layer coding method was used for encoding and decoding;the global machining route was generated uniformly when the algorithm initialized the candidate solutions;self-crossover was integrated into the search process to enhance local search;variable neighborhood mutation was introduced to avoid falling into local optimum;variable process number elite retention strategy was designed in the iterative process to expand the search space.Finally,the model and algorithm were solved and analyzed by an example of a diesel locomotive production workshop,which verifies the effectiveness of the model and the superiority and applicability of the algorithm.