CN-122021281-A - Quick high-precision modeling optimization method and system for multistage serial centrifugal compressors

CN122021281ACN 122021281 ACN122021281 ACN 122021281ACN-122021281-A

Abstract

The invention discloses a rapid high-precision modeling optimization method and system for multistage serial centrifugal compressors, which comprises the steps of firstly defining a general parameter and target performance database of the multistage compressors, completing global optimization of air flow angles based on a particle swarm optimization algorithm by taking the inverse of the square sum of errors of predicted values and target values of outlet pressures/temperatures of the compressors as fitness functions, outputting optimal air flow angle parameters, rapidly establishing a compressor model step by taking mass flow conservation as core constraint, completing speed decomposition, theoretical work calculation, air flow loss and thermodynamic parameter output of each stage, and finally coupling an air storage chamber model based on the mass conservation and the energy conservation to form an integrated modeling scheme of air flow angle optimization-multistage pneumatic modeling-air storage chamber dynamic coupling through absolute error and relative error verification model precision. The invention realizes the deep fusion of multistage compressor modeling and airflow angle optimization and the dynamic characteristics of the air storage chamber, and remarkably improves the precision compared with the traditional empirical modeling.

Inventors

ZHANG XIANLING
WU XIAOSHUANG
SHI SHAOLIANG
CHEN HANWEN
QIN CHAOHUI
CAO YUWEI
ZHANG BAOJUN
WEI SHANGSHANG
PANG LIBO
LI YIGUO
ZHOU JIAGEN
DENG JIANJUN

Assignees

华能中盐(常州)储能有限公司
华能国际电力江苏能源开发有限公司
东南大学

Dates

Publication Date: 20260512
Application Date: 20260121

Claims (8)

1. A rapid high-precision modeling optimization method for multistage serial centrifugal compressors is characterized by comprising the following steps: S1, defining general parameters and target performance data of a multistage compressor, wherein the general parameters comprise compression stages, gas physical parameters, mechanical efficiency and fixed mass flow, and the target performance data comprises outlet pressure, outlet temperature, power consumption and internal efficiency of each stage; S2, configuring airflow angle optimization parameters based on a particle swarm optimization algorithm, and setting constraint ranges of airflow angles at all levels according to the aerodynamic design theory of impellers and diffusers so as to avoid airflow separation or surge risks; S3, initializing the position and the speed of a particle swarm, taking the inverse of the error square sum of the predicted value and the target value of the outlet pressure and the temperature of the compressor as an fitness function, iteratively updating the individual optimal position and the global optimal position of the particles until convergence, and outputting the optimal airflow angle parameter; S4, establishing a four-stage compressor dynamic model step by step based on the optimal airflow angle parameters, and calculating the speed, pressure, temperature, efficiency and power of inlets/outlets of each stage by taking mass flow conservation as constraint conditions to ensure the physical consistency of inter-stage parameter transmission; S5, establishing a dynamic model of the air storage chamber, calculating the change of the pressure and the temperature of the air storage chamber along with the inflation time based on the final outlet parameters of the compressor coupled with mass conservation and energy conservation, and realizing the parameter linkage of the compression process and the air storage process; S6, calculating absolute errors and relative errors of the optimized performance parameters and the target values of all levels, verifying model accuracy, outputting a global optimal air flow angle, the total performance of the compressor and the dynamic characteristics of the air storage chamber, and providing complete parameter support for engineering application.
2. The method of claim 1, wherein the fitness function is expressed as: Wherein the method comprises the steps of 、 Is the predicted value of the outlet pressure and the temperature of the s-th stage, 、 Is the s-th stage outlet pressure and temperature target value.
3. The method of claim 1, wherein the update formula for particle velocity and position is: Wherein the method comprises the steps of 、 For the speed and position of the ith particle at time t, w is the inertial weight, c 1 、c 2 is the learning factor, r 1 、r 2 is a [0,1] random number, For an individual optimal position of the i-th particle, Is the global optimum.
4. The method of claim 1, wherein the progressive computation of the four-stage compressor dynamic model comprises: The interstage parameter transmission, namely, the 1 st stage inlet pressure is ambient pressure, the inlet temperature is ambient temperature, the 2 nd-s stage inlet pressure inherits the upper stage outlet pressure, and the inlet temperature is the cooled temperature, so that the interstage parameter continuity is ensured; Calculating the radial speed, the circumferential speed, the radial speed and the circumferential speed of the outlet of each stage based on the optimal airflow angle and the geometric parameters of the impeller, and providing a basis for energy transfer analysis; Calculating theoretical work and loss, namely calculating theoretical work according to the speed parameter, and calculating total loss by combining the airflow friction loss, the diffusion loss, the secondary flow loss and the dynamic pressure loss to reflect the energy loss under the actual working condition; and outputting thermodynamic parameters, namely calculating the outlet temperature and outlet pressure of each stage based on theoretical work, total loss and gas physical parameters, and ensuring that the parameters accord with thermodynamic laws.
5. The method of claim 1, wherein the calculation of the dynamic model of the air reservoir is based on mass conservation and energy conservation, and specifically comprises: conservation of mass: Conservation of energy: Dynamic parameters: Wherein the method comprises the steps of , , 。
6. The method of claim 1, wherein the convergence condition in step S3 is that the iterative global optimum fitness varies by less than 50 consecutive iterations Stability and reliability of the optimized result are ensured.
7. The method of claim 1, wherein the relative error in step S6 is less than 5%, and the total performance of the compressor includes total power consumption, total pressure ratio, and pressure ratio of each stage, providing key parameters for system matching.
8. A multistage serial centrifugal compressor rapid high-precision modeling optimization system is characterized by comprising: the parameter definition module is used for defining general parameters and target performance data of the multistage compressor, wherein the general parameters comprise compression stages, gas physical parameters, mechanical efficiency and fixed mass flow, and the target performance data comprises outlet pressure, outlet temperature, power consumption and internal efficiency of each stage; the optimization parameter configuration module is used for configuring airflow angle optimization parameters based on a particle swarm optimization algorithm, and setting constraint ranges of airflow angles at all levels according to the pneumatic design theory of the impeller and the diffuser so as to avoid airflow separation or surge risks; the particle swarm optimization module is used for initializing the position and the speed of the particle swarm, taking the inverse of the error square sum of the predicted value and the target value of the outlet pressure and the temperature of the compressor as an fitness function, iteratively updating the individual optimal position and the global optimal position of the particles until convergence, and outputting the optimal airflow angle parameter; The dynamic model calculation module is used for establishing a four-stage compressor dynamic model step by step based on the optimal airflow angle parameters, calculating the speed, pressure, temperature, efficiency and power of inlets/outlets of each stage by taking mass flow conservation as constraint conditions, and ensuring the physical consistency of inter-stage parameter transmission; The air storage chamber model calculation module is used for establishing an air storage chamber dynamic model, calculating the change of the pressure and the temperature of the air storage chamber along with the inflation time based on the final-stage outlet parameters of the compressor coupled with mass conservation and energy conservation, and realizing the parameter linkage of the compression process and the air storage process; And the optimization calculation verification module is used for calculating absolute errors and relative errors of the optimized performance parameters and target values of all levels, verifying model accuracy, outputting a global optimal air flow angle, the total performance of the compressor and the dynamic characteristics of the air storage chamber, and providing complete parameter support for engineering application.

Description

Quick high-precision modeling optimization method and system for multistage serial centrifugal compressors Technical Field The invention relates to the field of core equipment design of advanced adiabatic compressed air energy storage (AA-CAES) systems, in particular to a rapid high-precision modeling optimization method and system for multistage serial centrifugal compressors. Background In an advanced adiabatic compressed air energy storage system, a multistage centrifugal compressor is core equipment for converting electric energy into compressed air, and the performance of the multistage centrifugal compressor directly determines the energy storage efficiency, the pressure lifting capacity and the operation stability of the system. However, the existing multistage compressor modeling technology has the following key technical drawbacks, and the requirement of the AA-CAES system on "fast, high-precision and engineering" modeling is difficult to meet: The interstage pneumatic coupling is lost, namely the inlet airflow angle (beta 1), the outlet airflow angle (beta 2) and the outlet airflow angle (alpha 2) of each stage of impeller directly influence the airflow speed decomposition and the energy transfer efficiency, the traditional modeling mostly adopts empirical value setting, and the coupling influence of the airflow angle on interstage pressure and temperature transfer is ignored, so that the simulation precision is low; under the condition of fixed mass flow, the density, the speed, the outlet pressure and the temperature of each stage of inlet are required to satisfy dynamic balance, and the current simplified model often ignores the interference of air flow friction loss and secondary flow loss on mass flow conservation, so that the modeling result has large deviation from the actual working condition; The dynamic coupling of the air storage chamber is lost, namely, after the air flow from the outlet of the compressor enters the air storage chamber, the pressure and the temperature are dynamically changed along with the inflation time and are required to be matched with the parameters of the outlet of the compressor in real time, and the traditional modeling mostly analyzes the fracture of the compressor and the air storage chamber, so that the energy linkage characteristic of the compression process and the air storage process cannot be reflected. The experimental dependence is strong, although a Particle Swarm Optimization (PSO) algorithm is used for single-parameter optimization in the prior art, an integrated scheme of 'airflow angle optimization-compressor performance modeling-air reservoir dynamic coupling' is not formed yet, and the traditional modeling depends on an actual machine experiment or an empirical formula, so that the problems of high experimental cost, poor precision under low rotating speed/variable working conditions and incapability of adapting to compressors with different geometric parameters exist, and the requirement of 'quick, high precision and engineering' of an AA-CAES system on compressor modeling is difficult to meet. Disclosure of Invention Aiming at the technical defects of low modeling speed, insufficient precision, no consideration of dynamic coupling of an air flow angle and an air storage chamber and low experimental economy of the existing multistage compressor, the main purpose of the invention is to provide a rapid high-precision modeling optimization method of multistage serial centrifugal compressors, which is suitable for accurate simulation of pneumatic performance of the multistage compressor, optimization of key air flow angle parameters and dynamic characteristic coupling analysis of the air storage chamber, and provides a modeling scheme capable of being directly applied for engineering design, performance prediction and risk control of compression side core equipment of an AA-CAES system. Another object of the invention is to propose a fast high-precision modeling optimization system for multistage tandem centrifugal compressors. The invention provides a rapid high-precision modeling optimization method for multistage serial centrifugal compressors, which has the core aims of realizing global optimal optimization of airflow angles (beta 1, beta 2 and alpha 2) of each stage of multistage compressors, solving the precision problem caused by traditional experience setting, accurately simulating the outlet pressure, temperature and efficiency of the compressors under the constraint of mass flow conservation, ensuring modeling results to meet actual working conditions, realizing real-time coupling analysis of dynamic characteristics of the compressors and air reservoirs, reflecting energy linkage of the whole process of compression-air storage and providing reliable simulation basis for AA-CAES system design. To achieve the above objective, an embodiment of a first aspect of the present invention provides a rapid high-precision modeling optimization meth