Dongkun Chen ^1* Qiushi Cui ² Dongdong Li ¹ Panqiu Ren ³

• ¹ Shanghai University of Electric Power, Shanghai, China
• ² Chongqing University, Chongqing, China
• ³ Dongfang Electric Group Dongfang Electric Motor Co. Ltd, Sichuan, China

This paper intends to provide key insights to the manufacturing industrial park designers for selecting the typical days of electric load and planning the resources for energy-producing infrastructure. First, a hybrid time-series model of energy-consuming equipment based on the autoregressive integral moving average model (ARIMA) and Temporal Convolutional Network (TCN) is generated. According to this model, the energy consumption (EC) curve of large equipment in the industrial park can be depicted. Moreover, the present study designed a TLSM-IPML (Typical Load Stratification Method for Industrial Parks with Manufacturing Load) algorithm based on the typical days selected method. The data clustering method is utilized to analyze the energy usage characteristics. Furthermore, an energy usagebased planning model is proposed, network constraints are considered, and a multi-optional method is designed to solve the problem. Finally, case studies validate the superior performance of TLSM-IPML in analyzing characters of energy consumption, and planning model in reducing MES (manufacturing industrial factory integrated energy system) economic costs. Keywords: Manufacturing industrial integrated energy system, load data clustering, typical day, load clustering, equipment modeling, integrated energy system planningThis section introduces the MES model with nonlinear power and gas flow equations, including the resource endowment assessment, renewable energy equipment models, and ESS model.Furthermore, a mathematical model of EC equipment is built to conduct EC analysis.In this section, models of the energy equipment are shown. The structure of IES contains the following parts: photovoltaic power generation model, wind power generation model, HP model, and so on.Solar energy is a clean and abundant resource. Estimating solar energy resources involves evaluating total solar radiation. This parameter is compared with standard data to determine solar energy levels, as shown in Table 1, total solar radiation, is the sum of direct and scattered radiation.