Discussion on using optimization technology to imp

Discussion on using optimization technology to improve the comprehensive performance of automotive engines

I. preface

since the 1990s, the competition in the automotive industry has shifted from a single performance competition to a multi-dimensional comprehensive competition in performance, environmental protection, energy conservation and so on. As far as automobile engines are concerned, in order to cope with the world energy crisis and reduce environmental pollution, their research and development work has focused on reducing fuel consumption, emissions, light weight and wear reduction. In these studies, optimization technology will be widely used

automotive engine and aeroengine belong to the category of thermal engine, and they have similarities in many aspects. In recent years, automotive engine optimization 3. Control accuracy: upper limit +2%, lower limit ⑴% has a certain foundation, while the optimization technology established and applied for aerospace engines has made great progress. Applying these advanced optimization techniques, especially multidisciplinary optimization techniques, to the optimal design of automobile engines is expected to improve the comprehensive performance of automobiles centered on energy conservation and environmental protection. The author gives a brief description of the current development of optimization technology in the field of automotive engines and aerospace engines, and discusses the potential of using optimization technology to improve the comprehensive performance of automotive engines

II. Research and application status of engine optimization technology

at present, the common focus of all kinds of engine research and development work includes reducing fuel consumption, emissions, mass and wear. In order to achieve these goals, the application of optimization technology in engine design is an important means. At present, the optimization of engine is mainly carried out in engine structure, materials, fuel and combustion, emissions and multidisciplinary optimization

(I) engine structure and material optimization technology

engine structure optimization is mainly to optimize the shape of key parts to improve engine performance. The research in this area includes: combining BP neural network and genetic algorithm to optimize the structure of aeroengine to obtain the optimal thrust weight ratio; By optimizing the structure of solid rocket motor, the lightest structural mass and the largest loading density can be obtained; The research status of structural optimization design of aeroengine blade disc at home and abroad is summarized, and a new method combining dynamic analysis with structural shape optimization design is proposed; This paper expounds the importance of cad/cfd technology in the design and development of automobile engine, and gives the application effect of cad/cfd technology in the design of intake manifold of EFI gasoline engine and helical airway of diesel engine; Combining the boundary element method with the penalty function optimization method, the shape optimization of the engine connecting rod under alternating tension and compression loads is studied; Based on an efficient finite element method, the optimal design of three-dimensional complex shape connecting rod is carried out; Based on finite element analysis and optimization technology, a structural optimization method of engine crankshaft is proposed; The optimization of rocket engine casing is carried out, and the optimization selection of stress ratio and pv/w is discussed

in order to improve the engine structure and make the engine lightweight, the optimization design of its materials is an important means. In recent years, new composite materials such as silicon carbide, silicon nitride, zirconia, graphite and synthetic graphite have been used in engine structures. By establishing the stress-strain analytical formula and the maximum stress criterion of each section of the engine composite blade, the maximum strength of the blade is optimized

the composite shell of solid rocket motor is optimized, so that the motor structure can obtain the minimum mass under the requirements of strength constraints

(II) engine combustion optimization technology

with the increasingly serious problems of energy and environmental pollution in the world, aircraft and cars, as major polluters and energy consumers, have attracted much attention. The combustion process of engine has a direct impact on energy conservation and environmental protection, and the research on the optimization of engine combustion process has been paid more and more attention

at present, it is mainly optimized from the aspects of injection system, intake manifold, combustion chamber shape and so on. In terms of engine combustion and injection system, with the help of advanced electronic control technology, it can accurately adjust the fuel supply, optimize the injection timing and injection times, control the mixing state in the cylinder and the fuel distribution in the combustion chamber, and reduce emission pollution. For the injection timing of the new pulsating electronically controlled fuel injection system, the optimization of engine direct injection technology is studied

multi-objective design method is adopted to optimize the matching of engine combustion system and valve train. In the optimization research of combustion process of new fuel engine, based on the establishment of the optimal control model of hydrogen fuel engine, a dual-mode control mode is proposed; The air-fuel ratio of natural gas vehicle engine is optimized by means of computer simulation analysis to improve the performance of the engine

(III) multidisciplinary optimization technology of engine

engine design is based on multiple disciplines such as structure, heat, combustion, intensity, vibration, fluid, heat transfer, etc., with many variable factors and large randomness. It is an optimization problem of variable mutual coupling system. Multidisciplinary design optimization obtains the overall optimal solution of the system by making full use of the synergy generated by the interaction between various disciplines, so it has great application advantages in the traditional design flow chart of engine design figure 1

in the field of aero-engine, multidisciplinary optimization technology has been used to establish optimization models and carry out turbine blade design, compressor blade design and engine overall scheme design. The traditional optimization design method (as shown in Figure 1) is transformed into the multidisciplinary optimization parallel design process shown in Figure 2, which comprehensively considers the requirements of aerodynamics, vibration, strength and fatigue life, and can shorten the design cycle and improve the optimization effect. For example, the multidisciplinary optimization design of aeroengine nozzle is carried out by using single-stage optimization algorithm; Multidisciplinary robust design optimization method is introduced into the optimization research of internal combustion engine to evaluate the uncertainty in the design process; Multidisciplinary optimization technology is used to design the engine power system

III. research content of engine multidisciplinary multi-objective optimization technology

(I) multi-objective optimization

engine optimization involves multiple objectives. Unlike single objective optimization problems, these objective functions are often coupled together, and each objective has different physical significance and dimensions. Their relevance and conflict make it very difficult to optimize them. Multi objective optimization methods can be divided into the following two categories and have been applied in engine optimization design

1. preference based multi-objective optimization method

according to the actual situation of the project, this method first selects a preference vector, then uses the preference vector to construct a composite function, and uses the single objective optimization algorithm to optimize the composite function to find the optimal solution of a single protocol. For example, multi-objective optimization of engine mounting system is carried out by using linear combination method; The weight method is used to optimize the loss reduction and life extension control of liquid rocket engine

2. Multi objective optimization method based on non inferior solution set

this method first needs to find as many protocol solutions as possible, and then obtain the decision solution according to the actual situation of the project. Compared with the preference based multi-objective method, this method is more systematic, practical and objective. For example, through multi-objective genetic algorithm, the overall performance of aeroengine is optimized with unit thrust and fuel consumption as the objective function; The performance and cost of solid rocket motor are optimized based on multi-objective genetic algorithm

(II) uncertainty optimization

in the production and actual use of engines, there are always errors or uncertainties in material characteristics, manufacturing, assembly and load. Although in most cases, the error or uncertainty is small, the combination of these errors or uncertainties may have a great impact on the performance and reliability of the engine. For the optimization of such uncertain problems, the traditional optimization methods can not be solved, and we must resort to the uncertain optimization methods

1. random uncertainty optimization

in the random uncertainty optimization method, random variables or random functions are used to describe uncertainty, and the uncertain optimization problem is transformed into a deterministic optimization problem by using the methods of probability theory and mathematical statistics. The research of stochastic uncertain optimization is relatively mature, and it has been applied to engine design. For example, considering the randomness of geometry, load and boundary conditions, the structural reliability of aeroengine is studied; The probabilistic finite element stress analysis of cyclic symmetric structure in aeroengine is carried out

2. fuzzy uncertainty optimization

in the fuzzy uncertainty optimization method, fuzzy variables are used to describe the parameter uncertainty, and fuzzy statistical methods are used to study the uncertain phenomenon and convert the fuzzy optimization problem into a deterministic optimization problem for solution. For example, the optimization design of engine parts under multiple working conditions and multiple objectives

3. interval number uncertain optimization method

using intervals to describe the uncertainty of parameters, only need to know the upper and lower bounds of uncertain parameters, but do not need their accurate probability distribution and fuzzy membership function. Considering the convenience of this method, it may have a great application prospect in the future engine optimization design

(III) methods and Strategies of multidisciplinary optimization

the main idea of multidisciplinary optimization is to integrate the knowledge of various disciplines in the whole process of design, apply effective design optimization strategies and corresponding optimization methods, and organize and manage the design process. Its purpose is to obtain the overall optimal solution of the system by making full use of the synergy generated by the interaction between various disciplines. The main purposes of multidisciplinary optimization design are two: first, systematize the design process, that is, let the personnel of all disciplines involved in the design of the whole system understand the constraint requirements and optimization objectives of other disciplines, so as to avoid the design crash caused by mutual ignorance in the design process, which will lead to design changes and waste time and funds; Second, the design of each special feature runs through the whole process of design, that is, the traditional mechanical, electrical, control and other design majors should consider the comprehensive performance of reliability, maintainability, supportability, safety and testability in the process of realizing the traditional performance design. Multidisciplinary optimization methods mainly include single-level optimization, sequence optimization, parallel subspace optimization, collaborative optimization, multi-level hierarchical optimization and so on. At present, some advanced multidisciplinary optimization methods and strategies have been applied in the aerospace field. For example, on the iSIGHT software platform, the thrust and mass of an engine nozzle are optimized by using a single-stage optimization algorithm; Different multidisciplinary optimization methods and strategies are used for the parts, components and overall design of aeroengine

(IV) optimization algorithm

the optimization algorithm used in engine design includes both conventional optimization algorithm and intelligent optimization methods such as genetic algorithm and artificial neural network. Some current optimization algorithms are summarized and discussed, including the intergenerational mapping genetic algorithm with good global convergence performance and its improved method, the combination of conventional optimization algorithm and genetic algorithm, and the neural network with dynamic adjustment factors and dynamic learning rate

1. Conventional optimization techniques

such as simplex method, Newton method, feasible direction method, sequential quadratic programming, etc., these methods have mature theory, convenient application, and have been widely used