Authors: V. R. Vedruchenko, V. V. Krajnov, E. S. Lazarev /
Abstract
The subject of the study were made of the injection and combustion in diesel locomotives and ships, which are the most difficult to analyze, operate and forecasting facility management (processes in the cylinder of a diesel engine), where the conversion of thermal energy released during the combustion of fuel into mechanical work takes place through a series of successive physico-chemical, thermal, mass transfer and thermodynamic transformations which together constitute the circular irreversible and unlocked the duty cycle. However, the direct study of the operating cycle is still difficult complexity of the totality of the factors influencing the course of the process as a whole. The aim of the work was to establish an empirical connection laws fuel supply and combustion process in diesel cylinder it, ie the creation of methods of calculating these processes together. One approach to establishing formal relations and the possibility of subsequent numerical modeling of combustion processes and communication of the law the fuel in a diesel engine is the use of the theory of automatic control, which developed methods of identification, taking into account the processes of management system in the form of a model of a cybernetic system. It is concluded that the numerous experimental and computational studies suggest that the dynamics of the fuel significantly affects the combustion process in a diesel engine, and, consequently, its power and economic performance. Then to establish a formal link (model) combustion processes and the law the fuel used machine control theory (TAU), which uses and develops methods for the identification, considering processes management system in the form of a cybernetic circuit. Assumed that the full range of processes in the combustion chamber since the start of the fuel supply to the end of its combustion is a complex dynamic system of self-governing. The formulation was based on statistical methods for solving the problem of identification, where the input variable and its response (the law of supply and indicator diagram - experimental curves) are stationary random functions, and object management (combustion) is classified as a one-dimensional linear lumped. As a criterion of proximity to the object used criterion of the minimum of the expectation of a given function of the difference of the input signals and the object model. It is noted that there is currently no accurate method for calculating the finished analytically binding processes fuel supply, mixture formation and combustion in a diesel engine. Therefore, there is not yet calculated optimum injection characteristics suitable for the calculation and simulation processes in diesel engines of various types. It was found that, other things being equal, the law determined by the rate of injection of fuel injected. Striving to achieve injection with increasing speed in order to reduce the cycle dynamics, as well as more efficient use of the air charge in a distant "corners" of the combustion chamber (the latter portion of fuel, with a maximum speed to penetrate into the remotest corners). It is shown that in most cases it is necessary for the administration of the mathematical description of establishing the relationship between input and output variables, based on which can be worked out such a control object, which would ensure achievement of the intended target operation of the facility. With regard to the solution of a specific problem the most common case is when and exposure and response will be functions of the same argument. Thus, determining the experimental data of the conditional expectation of the output variable with respect to the input, we obtain the optimal (in the sense of the criterion of minimum mean square deviation) Estimates of the object. Given the accepted method of identification and characteristics of the problem, the best operator in the class of linear operators, and not among all possible random variables. This study was based on the principle of superposition, which is performed for a linear operator. Using the hypothesis that the investigated signals have the property of ergodicity with respect to the correlation functions. The adoption of the hypothesis of ergodicity possible to determine the correlation functions centered on the realization of a random process, ie, combustion process, as successive cycles. Noted that most fully identifiable objects are described in terms of the state space. Under the state of the object understood set of values, fully define its position at any given time. Proposed a model of the considered dynamic objects to choose a system of differential equations. In view of the above, the proposed approximation algorithm to determine the dynamic characteristics of the combustion process, ie, the agent object in the class of linear operators, which was presented as a system of differential equations representing the desired mathematical model.
