The dissertation is devoted to a new approach to solving the problem of grain drying, which is to create an adaptive optimal system of automatic control of the drying process, which allows to obtain the conditioned grain for quality further storage.
In the first section an analysis of grain properties as an object of drying is carried out, experimental regularities of the grain drying process, modes and methods of drying are considered.
According to the results of the analysis in the first section, it was established that the heat resistance of the grain is characterized by the maximum temperature of heating, which ensures complete storage of all qualitative indicators of grain. The limit of the zero degree of denaturation of proteins is the basis for choosing the grain drying regime.
It is established that during drying of preheated grain, the surface gradient of moisture content is less than that of conventional convective drying, that is, the evaporation zone is located near the outer surface of the material to which the moisture moves in the form of a liquid. This prevents overheating of the surface of the material and promotes the storage and improvement of the quality of the grain.
As a result of the analysis of grain properties and drying processes, as a compromise on quality and cost, it is advisable to take as a basis for the creation of a new mini-dryer a mine direct-feed grain dryer.
In the second section it is established that drying of grain in a mine direct-feed grain dryer is described by a mathematical model with spatial distribution of parameters.
A study of a static dryer model is realized by solving static equations with allowance for boundary conditions. In this case, the solution of the static model as a smooth spatial coordinate function determines the smooth distribution of the defining parameter along spatial coordinates. This allows to limit the number of primary transducers to measure the temperature and humidity of the grain in the dryer shaft.
Proceeding from the physical foundations of the drying process and the structure of the mathematical model of a dryer with three-dimensional distribution in the parameter space in the Cartesian coordinate system with boundary conditions of the third kind and the smoothness of the spatial temperature distribution (and humidity respectively) of the grain mass, in order to simplify the construction of the system of automatic process control, it is expedient to move from a system with distributed to a system with lumped parameters, introducing it to the matrix transmitted communication function i-th inputs with j-th output variables, which are measured in corresponding points of the unit with a limited number of primary temperature and humidity converters.
In the third section, the structure and algorithm of functioning of a multidimensional autonomous ACS with a reference model, which fulfills the task of drying the grain with the necessary quality, is developed.
The analysis of the simulation schedules showed that the developed control system satisfies the quality indicators at the determined parameters of the control object. However, during the drying process, the grain parameters change. This leads to the fact that the system cannot manage the drying process qualitatively (there is overregulation, the time of the transient process increases) or it becomes unmanageable at all. Therefore, in order to solve this problem, it is necessary to monitor the change of the parameters of the control object. This task is solved by building a subsystem of identification. The complement of the system of autonomous optimal (in the sense of conformity with the standard) control of the dryer as a multidimensional object, the subsystem of identification of the parameters of direct and cross-links in the control object, allows us to create a functionally reliable ACS in conditions of unsteadiness and stochastic characteristics of the object.
In the fourth section, using the system approach and the structure and algorithms of the operation of the multiple-adaptive identification system, a subsystem of process identification in the dryer was built.
Taking into account the standard situation for the use of the least squares method for the real situation of noise in both the output and the input variables of the identification object, the method of minimizing the optimally weighed amount was tested and used more perfectly (in the sense of unboundness and efficiency of the parameter estimates) symmetric correlation functions.
As a result of the use of the ILSM, unmatched and effective estimates of the transfer functions of the dryer have been obtained, which allowed to implement optimal adaptive algorithms for autonomous control of separate channels of humidity temperature of the dryer as a multidimensional object with cross-links.
