Optimization Methods and Systems for Combined Cycle Unit Operation
By optimizing the operation mode of the combined cycle unit using radial basis neural network and artificial fish swarm algorithm, the problem of error accumulation was solved, and efficient optimization of the LM6000 gas-steam combined cycle unit was achieved, reducing energy consumption and improving the stability of grid dispatch response, and supporting the localization of gas turbines.
Patent Information
- Application Number
- CN202410463866.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-17
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2044-04-17
Smart Images

Figure CN118375499B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of combined cycle unit operation technology, specifically to a method and system for optimizing the operation mode of combined cycle units, and more specifically to a method for optimizing the operation mode of combined cycle units based on radial basis function neural networks and fish swarm algorithm. Background Technology
[0002] Current research on the operation modes of combined cycle power units adopts the method of collecting basic data from on-site operation, and then analyzing the impact of errors in each component to output overall operating parameters. However, this method uses a simple accumulation algorithm, and the accumulation of errors in each part will lead to an increase in the overall error. Moreover, it does not analyze the relationship between input and output parameters, and the use of only a simple algorithm results in a large error.
[0003] Patent document CN112627921B (application number: 202011394886.4) discloses a gas... The method for optimizing the operation of combined cycle steam units, belonging to the field of gas turbine technology, includes the following steps: Step 1: [The text abruptly ends here, likely due to an incomplete sentence or missing information.] Step 1: Data from the distributed control system of the combined cycle steam turbine unit is mined and processed to summarize effective operational data; Step 2: Based on gas... The calculation relationships and mechanisms of the operating parameters of the combined cycle steam unit are decomposed layer by layer, and the effective data is classified and processed. Step 3: Combine the classified effective data and use the principles of performance calculation and consumption difference analysis to perform calculations. Step 4: Draw the optimal operating mode diagram or curve based on the calculation results. Step 5: Feed back to the operation center through the information network to assist in optimizing the unit's operation.
[0004] Patent document CN116163815A (application number: 202310065809.1) discloses a performance optimization method for a three-pressure reheat combined cycle power plant system, belonging to the field of gas turbine combined cycle power generation technology. The method includes the following steps: obtaining the key equipment parameter constraints used in the combined cycle power plant; obtaining the key parameters and margins for gas turbine thermal regulation; determining the operating envelope of the combined cycle power plant; establishing a thermal balance model for the gas turbine and the combined cycle power plant, calibrating the compressor and turbine efficiencies and the high-pressure cylinder efficiency; defining the matching strategy and optimization boundary for the main steam temperature and reheat steam temperature; establishing a combined cycle prior performance monitoring system to access real-time operating key parameters and environmental conditions; using the thermal balance model to traverse and calculate the combined cycle thermal efficiency under different parameter combinations to obtain the optimal parameters; feeding the optimal parameters back to the control system, and adjusting the system to achieve optimal unit performance. This patent only optimizes the performance of steam thermal energy through thermodynamic analysis, without taking into account the characteristics of combined cycle units. Although it improves thermal efficiency, it does not effectively analyze the gas turbine, the main equipment in the combined cycle unit, and has little guiding significance for the operation of the combined cycle unit.
[0005] My master's thesis, "Performance Analysis and Optimization of Gas-Steam Combined Cycle Units," published in June 2016, conducted performance analysis and multi-objective optimization research on gas-steam combined cycle units under varying operating conditions. It discussed the optimal allocation of thermal and electrical loads among different types of combined cycle units, established analytical calculation models for the main components of the gas-steam combined cycle unit, obtained the loss distribution under the unit's design and operating conditions, and performed variable operating condition analysis on the unit under different gas turbine loads, ambient temperatures, and heating capacities, obtaining the entropy variation laws of the unit and its components under different operating conditions. Secondly, based on economic theory, a cost model for the combined cycle unit was established. With the objectives of maximizing plant efficiency, minimizing total cost, and minimizing CO2 emissions, the NSGAⅡ genetic algorithm was used to conduct multi-objective optimization research on the combined cycle unit, obtaining the optimal parameter solution set, and then selecting the optimal parameter solution according to requirements. Furthermore, a method for optimizing the heat and power load allocation in gas-steam combined cycle units was proposed. Energy consumption characteristic curves for one-to-one and two-to-one combined cycle units were established for different types of units to optimize heat and power load allocation and obtain the best solution, providing a reference for power plant personnel. Finally, performance monitoring and operation optimization software for gas-steam combined cycle units was developed, and the software architecture, development process, and related functions were described in detail. However, this method performs a holistic analysis of the gas turbine from an equipment perspective, and the model uses a heavy-duty gas turbine, while the light-duty LM6000 gas turbine operates differently, resulting in significant errors. Summary of the Invention
[0006] To address the shortcomings of existing technologies, the purpose of this invention is to provide a method and system for optimizing the operation of combined cycle power units.
[0007] A method for optimizing the operation mode of a combined cycle power unit according to the present invention includes: Step S1: Perform thermodynamic and economic analysis of gas-steam combined cycle for gas turbine, waste heat boiler, steam turbine and combined cycle unit respectively, and obtain the factors affecting the variable operating characteristics of gas-steam combined cycle unit; Step S2: Based on the obtained factors affecting the variable operating characteristics of the gas-steam combined cycle unit, establish a variable operating characteristic prediction model of the gas-steam combined cycle unit based on the radial basis neural network, and train the constructed variable operating characteristic prediction model of the gas-steam combined cycle unit. Step S3: Based on the trained gas-steam combined cycle unit variable operating characteristic prediction model, the artificial fish swarm algorithm is used to obtain the optimal operating scheme of the combined cycle unit.
[0008] Preferably, step S1 involves: conducting thermodynamic and economic analyses of the gas-steam combined cycle for the gas turbine, waste heat boiler, steam turbine, and combined cycle unit respectively; constructing an objective function for economic optimization with the minimum natural gas consumption through a mathematical model of power plant revenue; and obtaining the variables affecting the objective function for economic optimization of the gas-steam combined cycle unit. The variables affecting the economic optimization objective function of the gas-steam combined cycle unit include: unit power, unit heat supply, compressor inlet temperature, turbine back pressure, and waste heat boiler flue gas temperature.
[0009] Preferably, step S2 employs: Step S2.1: Based on the obtained factors affecting the variable operating characteristics of the gas-steam combined cycle unit, establish a predictive model for the variable operating characteristics of the gas-steam combined cycle unit based on the radial basis neural network; Step S2.2: Select variables based on historical data of the combined cycle unit, including: unit power, unit heat supply, compressor inlet temperature, turbine back pressure, waste heat boiler flue gas temperature, and gas flow rate; Step S2.3: Normalize the selected variable data to obtain normalized variable data; Step S2.4: Use the gradient descent method to train the variable characteristic prediction model of the gas-steam combined cycle unit using the normalized variable data.
[0010] Preferably, step S2.1 involves constructing a predictive model for the variable operating characteristics of a gas-steam combined cycle unit by using the unit power, unit heating capacity, compressor inlet temperature, turbine back pressure, and waste heat boiler flue gas temperature as input variables of a radial basis neural network, and using gas flow rate as the output.
[0011] Preferably, step S3 employs the following methods: Step S3.1: Based on the different influences of artificial fish foraging behavior, swarming behavior, tail-chasing behavior and random behavior on the optimal value of the objective function and its position variables in the full space search within the simulated field of view of artificial fish, different parameters are set to achieve the purpose of global optimization. Step S3.2: Perform parameter analysis on artificial fish population size, global field of view, swimming step length, number of optimization attempts, and crowding factor. Based on the impact of parameter changes on convergence, determine the parameter settings required for the artificial fish swarm algorithm of the model. Step S3.3: The neural network-based model provides the objective variables of the economic optimization objective function for the gas-steam combined cycle unit, including: unit power or unit heat supply. The optimal solution of the objective variables, the minimum natural gas consumption, is obtained through the artificial fish swarm algorithm.
[0012] A combined cycle unit operation mode optimization system according to the present invention includes: Module M1: Perform thermodynamic and economic analysis of gas-steam combined cycle units for gas turbines, waste heat boilers, steam turbines, and combined cycle units respectively, and obtain the factors affecting the variable operating characteristics of gas-steam combined cycle units; Module M2: Based on the acquired factors affecting the variable operating characteristics of the gas-steam combined cycle unit, a variable operating characteristic prediction model for the gas-steam combined cycle unit is established based on the radial basis neural network, and the constructed variable operating characteristic prediction model for the gas-steam combined cycle unit is trained. Module M3: Based on the trained gas-steam combined cycle unit variable operating characteristic prediction model, the artificial fish swarm algorithm is used to obtain the optimal operating scheme of the combined cycle unit.
[0013] Preferably, module M1 employs the following methods: conducting thermodynamic and economic analyses of gas-steam combined cycle units for gas turbines, waste heat boilers, steam turbines, and combined cycle units respectively; constructing an objective function for economic optimization with the minimum natural gas consumption through a mathematical model of power plant revenue; and obtaining the variables affecting the objective function for economic optimization of gas-steam combined cycle units. The variables affecting the economic optimization objective function of the gas-steam combined cycle unit include: unit power, unit heat supply, compressor inlet temperature, turbine back pressure, and waste heat boiler flue gas temperature.
[0014] Preferably, the module M2 adopts: Module M2.1: Based on the acquired factors affecting the variable operating characteristics of gas-steam combined cycle units, a predictive model for the variable operating characteristics of gas-steam combined cycle units is established using a radial basis function neural network. Module M2.2: Based on historical data of combined cycle units, the variables selected include: unit power, unit heat supply, compressor inlet temperature, turbine back pressure, waste heat boiler flue gas temperature, and gas flow rate; Module M2.3: Normalizes the selected variable data to obtain normalized variable data; Module M2.4: Using normalized variable data, the gradient descent method is used to train the variable operating characteristic prediction model of the gas-steam combined cycle unit.
[0015] Preferably, module M2.1 adopts the following approach: using unit power, unit heating capacity, compressor inlet temperature, turbine back pressure, and waste heat boiler flue gas temperature as input variables of a radial basis neural network, and gas flow rate as the output quantity, to construct a predictive model for the variable operating characteristics of a gas-steam combined cycle unit.
[0016] Preferably, the module M3 adopts: Module M3.1: Based on the different effects of artificial fish foraging behavior, clustering behavior, tail-chasing behavior, and random behavior on the optimal value of the objective function and its positional variables in the full space search within the simulated field of view, different parameters are set to achieve the purpose of global optimization. Module M3.2: Parameter analysis is performed on the artificial fish population size, global field of view, swimming stride, number of optimization attempts, and crowding factor. Based on the impact of parameter changes on convergence, the parameter settings of the artificial fish swarm algorithm required for the model are determined. Module M3.3: The neural network-based model provides the objective variables for the economic optimization objective function of gas-steam combined cycle units, including: unit power or unit heat supply. The optimal solution for the objective variables, the minimum natural gas consumption, is obtained through the artificial fish swarm algorithm.
[0017] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention uses artificial neural network learning to model the variable operating conditions of the LM6000 gas-steam combined cycle unit, and artificial fish swarm algorithm to optimize the operation mode of the combined cycle unit. It formulates a scientific and economical load allocation method / means, which solves the problems of high energy consumption per unit of power generation and slow response to grid dispatch in combined cycle power generation enterprises, and achieves the effect of reducing energy consumption per unit of power generation and responding to grid dispatch efficiently and stably. 2. This invention achieves a breakthrough in the research of the operation process of the LM6000 gas-steam combined cycle unit by modeling and optimizing the operation process of the LM6000 gas-steam combined cycle unit. It combines neural network learning, artificial fish swarm algorithm and the research of the operation process of the gas-steam combined cycle unit, realizes the combination of theoretical research and production practice, and improves the generalizability of theoretical research. 3. Through the study of operating modes, this invention provides assistance in the mechanical research of equipment, offers strong support for the development of gas turbine operating technology in my country, and provides new ideas and directions for the localization and promotion of combined cycle units; 4. This invention, through the combination of neural network modeling and artificial fish swarm algorithm, realizes a new approach to the research of gas-steam combined cycle units, provides efficient operation guidance for tedious operation work, and improves the efficiency and effectiveness of on-site operation. Attached Figure Description
[0018] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 Flowchart of the method for optimizing the operation mode of combined cycle units.
[0019] Figure 2 A schematic diagram of a system for optimizing the operation mode of a combined cycle unit.
[0020] Figure 3 The overall load curve is generated for modeling.
[0021] Figure 4 This is a diagram showing the optimized load distribution curve for the combined cycle unit.
[0022] Figure 5 This is a graph showing the optimized fuel consumption curve of the combined cycle unit. Detailed Implementation
[0023] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the protection scope of the present invention.
[0024] Example 1 This invention provides a method and system for optimizing the operation mode of a combined cycle unit. It employs a radial basis function neural network (RBN) approach to model the variable operating characteristics of an LM6000 gas-steam combined cycle unit, constructing a variable operating characteristic model for the gas-steam combined cycle unit. Through analysis of the power plant's revenue mathematical model, it establishes the minimum natural gas consumption as the objective function for load optimization and allocation. Finally, it uses an artificial fish swarm algorithm to solve the model, and the effectiveness and feasibility of the proposed method are verified through practical application.
[0025] The method for optimizing the operation mode of the combined cycle unit includes: A method for optimizing the operation mode of a combined cycle power unit according to the present invention includes: Step S1: Perform thermodynamic and economic analysis of gas-steam combined cycle for gas turbine, waste heat boiler, steam turbine and combined cycle unit respectively, and obtain the factors affecting the variable operating characteristics of gas-steam combined cycle unit; Step S2: Based on the obtained factors affecting the variable operating characteristics of the gas-steam combined cycle unit, establish a variable operating characteristic prediction model of the gas-steam combined cycle unit based on the radial basis neural network, and train the constructed variable operating characteristic prediction model of the gas-steam combined cycle unit. Step S3: Based on the trained gas-steam combined cycle unit variable operating characteristic prediction model, the optimal operating scheme of the combined cycle unit is obtained using the artificial fish swarm algorithm; Step S4: Apply the unit operation plan to actual production, guide the actual operation of the combined cycle unit, and make improvements in actual production. This can optimize the operation mode of the combined cycle unit, reduce natural gas consumption, and improve economic efficiency.
[0026] Specifically, step S1 uses a gas turbine, a waste heat boiler, and a steam turbine as different components of a combined cycle unit, where the economic efficiency of each component directly affects the overall economic efficiency of the combined cycle unit. The economic efficiency of the combined cycle unit serves as a key performance indicator, and changes in various influencing factors will affect the final result.
[0027] The economic analysis of gas turbines includes: (1) In a gas turbine, the following energy balance relationship is satisfied:
[0028] in, This represents the amount of heat carried by the air required to produce 1 kg / s of natural gas, expressed in kJ / s. This indicates the lower heating value of natural gas, expressed in kJ / s. This indicates the actual efficiency of the gas turbine combustion chamber; This indicates the power generation capacity of the gas turbine, expressed in kJ / s. This indicates the amount of heat carried by the natural gas leaking from the gas turbine to the outside, expressed in kJ / s. This represents the heat carried by the normal exhaust gas from a gas turbine, expressed in kJ / s. This indicates the amount of heat lost by the gas turbine through the air cooler, expressed in kJ / s.
[0029] In a gas turbine, the cycle efficiency can be expressed as:
[0030] From the thermodynamic cycle of the gas turbine, the following relationship can be obtained:
[0031] Therefore, we can conclude that:
[0032] In the formula: This indicates the mass of natural gas that 1 kg of air can support combustion, expressed in kg (natural gas) / kg (air). The specific work of the gas turbine is represented by the following two variables, which are calculated as follows:
[0033] In the formula: This represents the flow rate of natural gas entering the combustion chamber, expressed in kg / s. This refers to the flow rate of compressed air entering the combustion chamber from the compressor, expressed in kg / s. This indicates the power output of the gas turbine compressor; This indicates the energy loss during the air compressor's compression of air; , The calculations for these two variables are shown below:
[0034]
[0035] In the formula: , T1 and T2 represent the average isobaric specific heat capacity of natural gas and air, respectively, in kJ / kg·K; T3, in K, represents the inlet flue gas temperature of the natural gas turbine. The unit is K, which represents the temperature of natural gas after it undergoes isentropic expansion through a gas turbine; , They represent , The specific enthalpy of the corresponding natural gas, expressed in kJ / kg; This indicates the operating efficiency of the gas turbine compressor; This indicates the energy conversion efficiency of a gas turbine. The gas expansion ratio of a gas turbine can be determined by the gas pressure at the turbine inlet. , outlet exhaust pressure Calculated; This represents the compressor exhaust temperature after isentropic compression, in Kelvin (K). This indicates the air temperature at the gas turbine inlet, expressed in Kelvin (K). , They represent , The corresponding specific enthalpy of air, in kJ / kg; The compressor pressure ratio is indicated by the compressor discharge pressure. Import pressure Calculated.
[0036] The economic analysis of waste heat boilers includes: In a waste heat boiler without supplemental combustion, the energy balance relationship is as follows:
[0037] In the formula: This indicates the amount of heat carried by the circulating water as it enters the waste heat boiler, expressed in kJ / s. This represents the energy carried by steam during exhaust from the high-pressure cylinder of a steam turbine, expressed in kJ / s. This indicates the heat carried by the main steam, expressed in kJ / s. This indicates the heat carried by the reheat steam, expressed in kJ / s. This indicates the heat carried by medium-pressure steam, expressed in kJ / s. This represents the thermal energy carried by the low-pressure main steam, expressed in kJ / s. This indicates the amount of heat carried by the flue gas from the waste heat boiler, expressed in kJ / s.
[0038]
[0039] In actual calculations, the change in the specific heat capacity of the flue gas can be ignored, i.e., it is a constant value. Then the above formula can be transformed into:
[0040] In the formula: This indicates the temperature of the exhaust gas from the gas turbine, expressed in Kelvin (K). This indicates the temperature of the flue gas discharged after the exhaust gas undergoes another heat exchange in a waste heat boiler, and the unit is K. This indicates the temperature of the atmospheric environment, measured in Kelvin (K).
[0041] The economic analysis of steam turbines includes: In a steam turbine, the energy balance relationship is as follows:
[0042] At this point, the efficiency of the steam turbine can be expressed as:
[0043] In the formula, This indicates the active power output of the steam turbine, measured in kW.
[0044] The economic analysis of combined cycle units includes: Taking a two-to-one combined cycle unit as an example, the efficiency during combined operation can be expressed as:
[0045] In the formula: , These represent the output power of gas turbine No. 1 and gas turbine No. 2, respectively, in kW; 、 These represent the output heat of gas turbine No. 1 and gas turbine No. 2, respectively, in kJ / s; , This indicates the efficiency of gas turbine No. 1 and gas turbine No. 2; , This indicates the efficiency of waste heat boiler No. 1 and waste heat boiler No. 2; Indicates the operating efficiency of the steam turbine; , These represent the fuel quantities for gas turbine No. 1 and gas turbine No. 2, respectively, in kg / s.
[0046] From the above formula, it can be concluded that the thermal efficiency of a gas-steam combined cycle unit is related to the efficiencies of the two gas turbines, the two waste heat boilers, and the steam turbine itself.
[0047] In summary, the overall operating characteristics of a gas-steam combined cycle unit are related to the individual characteristics of the gas turbine, waste heat boiler, and steam turbine. Furthermore, when analyzing the overall operation of a combined cycle unit, the impact of each component on the overall characteristics must be considered. In the economic analysis of combined cycle units, the gas turbine's gas consumption rate per unit load and the overall plant power consumption rate are the main economic indicators. Moreover, in actual production, fuel consumption is the largest expense in unit operation; therefore, reducing the gas consumption rate and the plant power consumption rate are the most important approaches to economic analysis.
[0048] Based on the above thermodynamic and economic analysis of gas-steam combined cycle, it can be found that the economy of combined cycle units is related to various components such as gas turbine, waste heat boiler, and steam turbine. The air factors affecting the operating characteristics of gas-steam combined cycle units are mainly ambient temperature, ambient air pressure, and the temperature of compressed air entering the combustion chamber. Regarding ambient air pressure, atmospheric pressure changes are very small, only affecting the power consumption during the air compression process. Similarly, ambient temperature affects the state of the air but not its state before entering the combustion chamber.
[0049] For fuel combustion in the combustion chamber, the heat carried by the compressed air entering the combustion chamber affects the heat cycle of the gas turbine. Gas turbines and combined cycle units can clearly reflect the state of the air mixture during combustion. However, the specific heat capacity of the compressed air in the compressor changes relatively little and can be ignored. Considering the influence of existing power plant measurement data, the compressor inlet temperature (°C) is selected as the input variable for the combined cycle unit variable model.
[0050] After the high-temperature, high-pressure gas completes its work process in the turbine, it is transported to the waste heat boiler for reuse via the turbine's last-stage blades. For the waste heat boiler, the exhaust gas emitted from the turbine has excellent temperature and pressure. Without supplementary combustion, the exhaust gas only undergoes a heat exchange process in the waste heat boiler, without any supplementary heating process. At this time, the connection between the waste heat boiler and the outside world is mainly the temperature and flow rate of the flue gas discharged into the atmosphere after heat exchange.
[0051] Therefore, the main factor affecting the variable operating characteristics of a combined cycle unit is the state of the exhaust gas from the waste heat boiler. Based on the influence of the waste heat boiler measurement data, the impact of the waste heat boiler flue gas temperature on the variable operating characteristics of the unit is considered.
[0052] In a waste heat boiler without supplemental combustion, after heat exchange, the high-temperature, high-pressure exhaust gas heats water to produce high-pressure, high-temperature steam. This steam enters the turbine to perform work, converting thermal energy into mechanical energy. Simultaneously, the exhaust steam is discharged to the condenser for cooling and condensation. Pressurized by the feedwater pump, it enters the next cycle. In the cooler, the pressure level (i.e., the turbine back pressure (MPa)) significantly affects the characteristics of the combined cycle unit. The main factors influencing the pressure level are the cooling medium, specifically the temperature and flow rate of the circulating cooling water. Conversely, the temperature and flow rate of the power plant's circulating cooling water affect the characteristics of the combined cycle unit, which are reflected in the condenser pressure.
[0053] In summary, when considering the influence of cooling medium factors, the change in turbine back pressure (MPa) is selected as the input variable for the combined cycle unit variable operating characteristic model.
[0054] For a gas-steam combined cycle unit, the energy conversion is as follows: First, the high-temperature, high-pressure gas does work on the gas turbine, converting the gas's thermal energy into the mechanical energy of the gas turbine. The gas turbine then drives the generator set, converting the mechanical energy into electrical energy. Second, the high-temperature, high-pressure steam generated after heat exchange in the waste heat boiler enters the steam turbine, converting its kinetic energy into mechanical energy. The steam turbine then drives the generator set, converting the mechanical energy into electrical energy. Both energy conversion paths are related to the generator set's power and the unit's heat supply. Furthermore, the generator set's power conversion is the primary pathway for gas thermal energy conversion. Therefore, in studying the variable operating characteristics of a gas-steam combined cycle unit, the influence of generator set power and the unit's heat supply needs to be considered.
[0055] Specifically, step S2 employs the following: Step S2.1: Based on the obtained factors affecting the variable operating characteristics of the gas-steam combined cycle unit, establish a predictive model for the variable operating characteristics of the gas-steam combined cycle unit based on the radial basis neural network; Step S2.2: Select variables based on historical data of the combined cycle unit, including: unit power, unit heat supply, compressor inlet temperature, turbine back pressure, waste heat boiler flue gas temperature, and gas flow rate; Step S2.3: Normalize the selected variable data to obtain normalized variable data; Step S2.4: Use the gradient descent method to train the variable characteristic prediction model of the gas-steam combined cycle unit using the normalized variable data.
[0056] Step S2.2 adopts the following approach: selecting unit power (MW), unit heat supply, compressor inlet temperature (°C), turbine back pressure (kPa), and waste heat boiler flue gas temperature (°C) as input variables for radial basis function neural network modeling, and using gas flow rate (Nm2) as output quantity, and then using radial basis function neural network for modeling. Step S2.3 adopts the following approach: In order to improve the accuracy of the modeling of the combined cycle unit under different operating conditions, it is necessary to convert and limit the training data and test data to a certain range. In the radial basis neural network toolbox of the MATLAB simulation software, the most accurate method is when the calling function requires the input variables to be data between 0 and 1.
[0057] Specifically, step S3 employs the following: Step S3.1: Based on the different influences of artificial fish foraging behavior, swarming behavior, tail-chasing behavior and random behavior on the optimal value of the objective function and its position variables in the full space search within the simulated field of view of artificial fish, different parameters are set to achieve the purpose of global optimization. Step S3.2: Perform parameter analysis on artificial fish population size, global field of view, swimming step length, number of optimization attempts, and crowding factor. Based on the impact of parameter changes on convergence, determine the parameter settings required for the artificial fish swarm algorithm of the model. Step S3.3: The model based on the neural network algorithm provides the objective variables of the economic optimization objective function of the gas-steam combined cycle unit, and the optimal solution of the objective variables is obtained by the artificial fish swarm algorithm.
[0058] The objective function is to minimize the natural gas consumption per unit of power generation of the combined cycle unit, and the load output of the gas turbine unit is the decision variable. The solution process of the combined cycle unit scheduling optimization model based on the artificial fish swarm algorithm is as follows: 1) visual=100; step=1; try_number=10; delta=35; fishnum=50; Set the total number of simulated artificial fish groups to 50, the crowding factor to 35, the maximum number of iterations to 50, the swimming step size to 1, the field of view radius to 100, and the maximum number of trials to 10.
[0059] 2) In the initial state of the fish swarm, assign the minimum natural gas consumption per unit of power generation to the bulletin board.
[0060] 3) After artificial fish in simulated states exhibit foraging behavior, grouping behavior, and tail-chasing behavior, the state vectors before and after the behavior are compared, and the optimal behavior is selected for execution. The state of an individual is evaluated by the amount of natural gas consumed per unit of electricity generated.
[0061] 4) Bulletin Board Status Update Rules: By comparing fitness values, if the natural gas consumption per unit of power generation on the bulletin board is better than the status of the individual fish, it remains unchanged; if the individual fish is better than the status on the bulletin board, that is, the natural gas consumption per unit of power generation of the individual fish is better than the status on the bulletin board, the bulletin board status is updated.
[0062] 5) Determine the output value by judging the number of iterations. If the bulletin board status has reached the maximum number of iterations, output the variable value with the minimum natural gas consumption per unit of power generation; otherwise, jump to step 3.
[0063] The present invention also provides a combined cycle unit operation mode optimization system, which can be implemented by executing the process steps of the combined cycle unit operation mode optimization method. That is, those skilled in the art can understand the combined cycle unit operation mode optimization method as a preferred embodiment of the combined cycle unit operation mode optimization system.
[0064] Those skilled in the art will understand that, besides implementing the system and its various devices, modules, and units provided by this invention in the form of purely computer-readable program code, the same functions can be achieved entirely through logical programming of the method steps, making the system and its various devices, modules, and units of this invention function in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, the system and its various devices, modules, and units provided by this invention can be considered as a hardware component, and the devices, modules, and units included therein for implementing various functions can also be considered as structures within the hardware component; alternatively, the devices, modules, and units for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.
[0065] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.
Claims
1. A method for optimizing the operation mode of a combined cycle power unit, characterized in that, include: Step S1: Perform thermodynamic and economic analysis of gas-steam combined cycle for gas turbine, waste heat boiler, steam turbine and combined cycle unit respectively, and obtain the factors affecting the variable operating characteristics of gas-steam combined cycle unit; Step S2: Based on the obtained factors affecting the variable operating characteristics of the gas-steam combined cycle unit, establish a variable operating characteristic prediction model of the gas-steam combined cycle unit based on the radial basis neural network, and train the constructed variable operating characteristic prediction model of the gas-steam combined cycle unit. Step S3: Based on the trained gas-steam combined cycle unit variable operating characteristic prediction model, the optimal operating scheme of the combined cycle unit is obtained using the artificial fish swarm algorithm; Step S1 involves: conducting thermodynamic and economic analyses of gas-steam combined cycle units for gas turbines, waste heat boilers, steam turbines, and combined cycle units respectively; constructing an objective function for economic optimization with the minimum natural gas consumption through a mathematical model of power plant revenue; and obtaining the variables affecting the objective function for economic optimization of gas-steam combined cycle units. The variables affecting the economic optimization objective function of the gas-steam combined cycle unit include: unit power, unit heat supply, compressor inlet temperature, turbine back pressure, and waste heat boiler flue gas temperature. Step S2 employs the following: Step S2.1: Based on the obtained factors affecting the variable operating characteristics of the gas-steam combined cycle unit, establish a predictive model for the variable operating characteristics of the gas-steam combined cycle unit based on the radial basis neural network; Step S2.2: Select variables based on historical data of the combined cycle unit, including: unit power, unit heat supply, compressor inlet temperature, turbine back pressure, waste heat boiler flue gas temperature, and gas flow rate; Step S2.3: Normalize the selected variable data to obtain normalized variable data; Step S2.4: Use the gradient descent method to train the prediction model of the variable operating characteristics of the gas-steam combined cycle unit using the normalized variable data; Step S2.1 adopts the following approach: using unit power, unit heat supply, compressor inlet temperature, turbine back pressure, and waste heat boiler flue gas temperature as input variables of a radial basis neural network, and gas flow rate as the output quantity, to construct a predictive model for the variable operating characteristics of a gas-steam combined cycle unit. Step S3 employs the following: Step S3.1: Based on the different influences of artificial fish foraging behavior, swarming behavior, tail-chasing behavior and random behavior on the optimal value of the objective function and its position variables in the full space search within the simulated field of view of artificial fish, different parameters are set to achieve the purpose of global optimization. Step S3.2: Perform parameter analysis on artificial fish population size, global field of view, swimming step length, number of optimization attempts, and crowding factor. Based on the impact of parameter changes on convergence, determine the parameter settings required for the artificial fish swarm algorithm of the model. Step S3.3: The neural network-based model provides the objective variables of the economic optimization objective function for the gas-steam combined cycle unit, including: unit power or unit heat supply. The optimal solution of the objective variables, the minimum natural gas consumption, is obtained through the artificial fish swarm algorithm.
2. A combined cycle unit operation mode optimization system, characterized in that, include: Module M1: Perform thermodynamic and economic analysis of gas-steam combined cycle units for gas turbines, waste heat boilers, steam turbines, and combined cycle units respectively, and obtain the factors affecting the variable operating characteristics of gas-steam combined cycle units; Module M2: Based on the acquired factors affecting the variable operating characteristics of the gas-steam combined cycle unit, a variable operating characteristic prediction model for the gas-steam combined cycle unit is established based on the radial basis neural network, and the constructed variable operating characteristic prediction model for the gas-steam combined cycle unit is trained. Module M3: Based on the trained gas-steam combined cycle unit variable operating characteristic prediction model, the artificial fish swarm algorithm is used to obtain the optimal operating scheme of the combined cycle unit; The module M1 employs the following approach: conducting thermodynamic and economic analyses of gas-steam combined cycle units for gas turbines, waste heat boilers, steam turbines, and combined cycle units respectively; constructing an objective function for economic optimization with the minimum natural gas consumption through a mathematical model of power plant revenue; and obtaining the variables affecting the objective function for economic optimization of gas-steam combined cycle units. The variables affecting the economic optimization objective function of the gas-steam combined cycle unit include: unit power, unit heat supply, compressor inlet temperature, turbine back pressure, and waste heat boiler flue gas temperature. The module M2 adopts: Module M2.1: Based on the acquired factors affecting the variable operating characteristics of gas-steam combined cycle units, a predictive model for the variable operating characteristics of gas-steam combined cycle units is established using a radial basis function neural network. Module M2.2: Based on historical data of combined cycle units, the variables selected include: unit power, unit heat supply, compressor inlet temperature, turbine back pressure, waste heat boiler flue gas temperature, and gas flow rate; Module M2.3: Normalizes the selected variable data to obtain normalized variable data; Module M2.4: Using normalized variable data, the gradient descent method is used to train the variable operating characteristic prediction model of the gas-steam combined cycle unit; The module M2.1 adopts the following approach: using unit power, unit heat supply, compressor inlet temperature, turbine back pressure, and waste heat boiler flue gas temperature as input variables of a radial basis neural network, and gas flow rate as the output quantity, to construct a predictive model for the variable operating characteristics of a gas-steam combined cycle unit. The module M3 adopts: Module M3.1: Based on the different effects of artificial fish foraging behavior, clustering behavior, tail-chasing behavior, and random behavior on the optimal value of the objective function and its positional variables in the full space search within the simulated field of view, different parameters are set to achieve the purpose of global optimization. Module M3.2: Parameter analysis is performed on the artificial fish population size, global field of view, swimming stride, number of optimization attempts, and crowding factor. Based on the impact of parameter changes on convergence, the parameter settings of the artificial fish swarm algorithm required for the model are determined. Module M3.3: The neural network-based model provides the objective variables for the economic optimization objective function of gas-steam combined cycle units, including: unit power or unit heat supply. The optimal solution for the objective variables, the minimum natural gas consumption, is obtained through the artificial fish swarm algorithm.
Citation Information
Patent Citations
A method for optimizing the operation of gas-steam combined cycle units
CN112627921B
Performance optimization method for three-pressure reheating combined cycle power plant system
CN116163815A
Operation optimization decision-making method for gas-steam combined cycle unit
CN112627921A
