The invention relates to an improved variable neighborhood search algorithm-based production scheduling method and system. The method comprises the steps of 1, grouping workpieces; 2, performing batching on the workpieces in each group; 3, setting a neighborhood structure of a solution; 4, setting a learning policy; 5, setting parameters in an algorithm; 6, generating an initial solution; 7, generating an initial population individual; 8, initializing an individual knowledge base and a social knowledge base; 9, performing updating to obtain a new solution of the individual; 10, calculating a fitness value of the individual; 11, updating the individual knowledge base and the social knowledge base; 12, obtaining an optimal solution in a population; 13, comparing the initial solution and theoptimal solution of the population, and if the initial solution is superior to the optimal solution of the population after updating, returning to the step 9; or otherwise, updating the initial solution, and returning to the step 7; and 14, judging whether a stop condition is met or not, and if the stop condition is met, ending the algorithm and outputting the optimal solution, or otherwise, continuing the algorithm. According to the method provided by the invention, the production efficiency can be improved and the production cost can be reduced, so that overall economic benefit maximizationof enterprises is realized.