An effective method for optimization of continuous and discrete parameters of heat and power plants

Modern thermal power plants are complex technological systems. Therefore, making informed decisions when studying them requires the use of mathematical modeling and nonlinear optimization methods for plant parameter. The most complex task is to solve a mixed optimization problem wherein a part of optimization parameters vary continuously, and the other can take only discrete (integer) values. An effective method is developed to solve a thermal power plant optimization problem with continuous and discrete parameters. The method suggests an iterative procedure for solving continuous nonlinear programming problems and discrete-continuous linear programming problems. For each iteration, we add new constraints obtained by linearizing nonlinear inequality constraints and the objective function of the initial problem to the system of inequality constraints of linear problem. The effectiveness of the proposed method is exemplified by the optimization of a combined cycle power plant with a mixture of working media (the STIG scheme). Design characteristics of heating surface of a waste heat recovery boiler represent the discrete parameters to be optimized. The research demonstrates a considerable reduction in computational effort compared to the branch and bound method.