The Energy Consumption Characteristic Analysis And Energy Efficiency Optimization Study Of Extrusion Manufacturing Considering Uncertainty

Extrusion manufacturing is a mechanical manufacturing method based on the plastic molding principle,with significant production and economic advantages.Therefore,it has attracted increasing attention.In this process,the metallic billet is processed into tube,rod,T–shaped,L–shaped and other profiles at one time by extrusion.This process is a high quality and efficient chip–free processing technology,which has achieved more and more applications in cutting–edge science and technology fields such as aerospace,rail transit,building decoration and national defense.Extruder is the most important equipment to realize extrusion manufacturing.With the rise of resource price and the increasingly fierce market competition,the research on its energy saving and consumption reduction becomes more and more urgent.This paper focuses on the scientific problem of extrusion manufacturing energy consumption characteristic analysis and energy efficiency optimization considering the uncertainty.Firstly,the research background and research status of the extrusion manufacturing process are reviewed.Then,from the process level,equipment level and system level,energy consumption characteristics analysis and energy efficiency optimization research are carried out.Finally,combined with the above research methods and theories,a set of low carbon energy saving extrusion manufacturing system was developed.Our study is done as follows:（1）At the process level,in order to solve the problem of dynamic energy consumption characteristic analysis and energy efficiency optimization of extrusion manufacturing,this thesis proposes an energy consumption characteristic analysis method based on Power Bond Graph.First,the basic energy flow of the extruder during the production process is analyzed.Then,the energy consumption characteristics of the hydraulic system are modeled based on the Power Bond Graph.Then,the equation of state for the energy consumption is derived.Then,the dynamic properties of the energy consumption were simulated using the MATLAB simulation software.Finally,the process parameters are optimized according to the proposed method,and the energy efficiency of extrusion manufacturing is optimized at the process level.（2）At the equipment level,in order to solve the dynamic real–time detection problem of extrusion manufacturing anomalous energy consumption,this thesis proposes a dynamic real–time unsupervised anomaly detection method.The proposed method provides a good analysis of the early energy consumption characteristics of the extrusion manufacturing equipment.As for point anomaly,two anomaly detection algorithms based on Rain Flow counting method are proposed and compare them.While,as for collective anomaly,a Rain Flow–based Mean Nearest Neighbor Distance Anomaly Factor algorithm is proposed,which is a distance–based detection method.An energy efficiency optimization framework based on anomaly detection is proposed.This provides an effective solution for clean production and low carbon manufacturing of extrusion manufacturing.（3）At the equipment level,in order to solve the problem of abnormal energy consumption diagnosis of long–term full load operation extruder under complex working conditions.This thesis presents two diagnostic models for anomalous energy consumption based on an improved Negative Selection Algorithm.This method provides a good analysis of the medium–term energy consumption characteristics of extrusion manufacturing equipment.A method combining the dendrogram with statistics-based feature selection is proposed,which aims to analyze the energy consumption influence factors of the extruder.In order to solve the problem that traditional Negative Selection Algorithms cannot judge abnormal energy consumption patterns,a new classifier generation method is proposed.The distribution of classifiers was optimized by ant colony and simulated annealing algorithms,aiming to improve diagnostic accuracy.The detection phase of the traditional Negative Selection Algorithm is improved using a directed acyclic graph to achieve the diagnosis of abnormal energy consumption.（4）At the equipment level,in order to solve the localization problem of energy consumption data with strong uncertainty,non–linear and non–Gaussian features,this thesis proposes an abnormal energy consumption localization technique considering uncertainty.The proposed method provides a good analysis of the late energy consumption characteristics of the extrusion manufacturing equipment.To quantitatively evaluate the correlation between abnormal machine components and energy consumption,an uncertainty evaluation method combining entropy weight and fuzzy theory is proposed,which aims to provide a decision basis for the location of abnormal energy consumption.To accurately locate abnormal energy consumption,a novel localization technique is proposed,which is called genetic algorithm–based wavelet neural network.Firstly,the energy consumption data is processed by wavelet packet decomposition and reconstruction.The energy value is calculated,and then the dimension reduction is performed by principal component analysis.（5）At the system level,to address the assessment of energy consumption in extrusion manufacturing systems,this thesis investigates a dynamically continuous,three–level,buffered–free hybrid manufacturing system driven by uncertain interrupt events.A real–time energy consumption modeling approach considering uncertain manufacturing systems is proposed,which aims to identify the energy consumption and energy losses of the manufacturing system in real time and predict potential energy losses in the future.This is an effective strategy based on data–driven real–time energy consumption assessment,which provides a decision basis for the energy efficiency optimization of manufacturing systems.（6）Based on the energy consumption characteristic analysis and energy efficiency optimization theory proposed in this thesis,a MATLAB GUI–based energy efficiency optimization system is developed.First,we analyze the requirements of the system,determine the overall framework,then design five functional modules,then introduce the development environment and workflow;finally,implement the optimization system,and introduce the operation method.