Cargo Volume Forecast and Personnel Scheduling Model of Logistics Network Based on ARIMA, LSTM and MOP

With the popularization of the Internet and the vigorous development of online shopping, the sorting center of logistics network is an important research topic in the logistics distribution link. Aiming at the problems of cargo volume and personnel scheduling in logistics network center, this paper establishes ARIMA model, LSTM model, BP neural network model and multi-objective programming model to solve the cargo volume and personnel scheduling in different sorting centers, and uses cluster analysis, GA algorithm and NSGA-II algorithm to solve the objectives. In this paper, the data is first preprocessed, including filling in missing values by linear interpolation, processing outliers, data visualization (box graph, 3σ graph), and standardization processing. Secondly, the stationarity of the time series is judged by the significant difference. If the stationarity is not stable, the stationarity of the time series is transformed by differential processing. Then the ARMI-LSTM model is established to predict the time series. In the personnel scheduling, taking the specific work scheduling and actual hourly efficiency of regular and temporary workers as the objective function, and taking the quantity limit and the number of attendance as the constraint conditions, a set of initial solutions are obtained through the multi-objective programming model, and then genetic algorithm is introduced to optimize them. Finally, a 0-1 linear programming model is established, and on this basis, the problem is solved by non-dominated sorting genetic algorithm.