A method and system for online correction of a nuclear reactor based on a model predictive control architecture

Abstract

The application discloses a digital twin architecture of a nuclear reactor, and relates to a nuclear reactor online correction method and system based on a model predictive control architecture, which comprises the following steps: S101, arranging various sensors such as neutron flux, temperature and control rod position in a reactor system; S201, obtaining the measurement values of the sensors in S101, such as reactor core power, reactor core inlet temperature and control rod position; S301, establishing a reactor digital twin model; S401, substituting the measurement values in S201 into the mathematical model in S301 as input, obtaining the output calculated by the digital twin model, and adjusting the adjustable parameters of the digital twin model to realize online correction. The application utilizes the idea of model predictive control, adjusts the adjustable parameters of the digital twin model according to the actual reactor state, thereby being capable of correcting the output of the digital twin model online, ensuring that the key state variables of the reactor under different working conditions are truly mapped, and enhancing the safety and reliability of the reactor.

Description

Technical Field

[0001] This invention belongs to the field of nuclear reactor digital twin technology, specifically, it relates to a method and system for online calibration of nuclear reactors based on a model predictive control architecture. Background Technology

[0002] Nuclear reactors have enormous application potential, but their operational stability in unattended environments is a crucial factor limiting the application of nuclear power. In remote mountainous areas, uninhabited islands, deep space exploration, and other unmanned / sparsely staffed environments, nuclear reactors face the problem of sparse communication data, making it difficult to accurately grasp their operational status. This leads to existing operating plans and control strategies failing to guarantee safe and stable system operation, resulting in unplanned shutdowns and loss of power. Realizing a digital twin of the nuclear reactor is an effective way to solve this problem. A digital twin can provide the physical manufacturing system with real-world information on its status and operation, enhancing the system's intelligence in analysis, evaluation, predictive diagnosis, and performance optimization.

[0003] Against this backdrop, a nuclear reactor online calibration method based on model predictive control architecture is proposed. By constructing a nuclear power digital companion model, the method achieves "virtual reflection of reality", fully reflecting the operating status of system parameters and enhancing the understanding of the operating status of nuclear power in unmanned / minimally manned conditions.

[0004] In view of this, the present invention is proposed. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide a method and system for online calibration of nuclear reactors based on model predictive control architecture, thereby solving the problems mentioned in the background art.

[0006] To solve the above-mentioned technical problems, the basic concept of the technical solution adopted by the present invention is as follows: An online calibration method for nuclear reactors based on a model predictive control architecture is proposed. This method employs model predictive control principles combined with intelligent optimization algorithms to optimize the adjustable parameters of a digital twin model, thereby achieving reactor state mapping. The method includes the following steps: S101. Neutron flux sensors, temperature sensors and control rod position sensors are installed at designated locations in the reactor system. Specifically, S101 includes: S1011. A neutron flux measurement sensor is installed inside the reactor pressure vessel to measure the neutron flux distribution in the reactor core in real time. S1012. A temperature sensor is installed on the surface of the reactor outlet nozzle to measure the temperature of the working fluid at the reactor core outlet in real time. S1013. Displacement sensors are arranged on the reactor control rod drive mechanism to measure the position of the control rods in real time. S201. Collect the measurement data from the sensor and process it to obtain the core power, core inlet working fluid temperature and control rod position parameters; Specifically, S201 includes: S2011. Obtain neutron flux data from various locations in the reactor core collected by the neutron flux measurement sensor, and calculate the reactor core power value accordingly. S2012. Obtain core inlet working fluid temperature data directly measured by temperature sensors; S2013. Obtain the control rod position data measured by the control rod position sensor; S301. Construct a digital twin model of the reactor, and screen and determine the adjustable parameters of the model; Specifically, S301 includes: S3011. Construct a reactor core point dynamics model; S3012. Construct a reactor thermal-hydraulic model that includes a fuel temperature calculation module and a coolant temperature calculation module; S3013. Evaluate the parameters of the constructed digital twin model of the nuclear reactor, preliminarily screen adjustable parameters, and determine the final adjustable parameters for online calibration through sensitivity analysis; S401. Input the parameters into the digital twin model to obtain the calculated output value, and adjust the adjustable parameters through model predictive control and intelligent optimization algorithm to complete the online correction; Specifically, S401 includes: S4011. Input the control rod position data obtained in S2013 into the reactor core point dynamics model constructed in S3011 to calculate the core power calculation value of the digital twin model. S4012. Input the core inlet working fluid temperature data obtained in S2012 into the reactor thermal-hydraulic model constructed in S3012 to calculate the core outlet working fluid temperature of the digital twin model. S4013. The calculated values ​​of core power and core outlet working fluid temperature are compared with the measured core power value in S2011 and the measured core outlet working fluid temperature value in S1012, respectively. Based on the model predictive control concept and combined with intelligent optimization algorithm, the calculation results of the digital twin model are adjusted to achieve online correction and map the reactor state.

[0007] A nuclear reactor online calibration system based on a model predictive control architecture, applying any of the online calibration methods described above, includes a data acquisition module, a reactor adjustable parameter reconfiguration module, and an online calibration module connected in sequence. The data acquisition module is used to acquire measurement data from sensors located throughout the reactor. The sensor measurements acquired by the data acquisition module include: core power measured by the neutron flux sensor, core outlet working fluid temperature measured by the temperature sensor, and control rod position measured by the control rod sensor. The reactor adjustable parameter reconfiguration module is used to screen adjustable parameters in the digital twin model, determine the final adjustable parameters through sensitivity analysis, and provide a parameter basis for online calibration. The reactor adjustable parameter reconfiguration module is used to screen adjustable parameters in the digital twin model. The adjustable parameters are affected by the core power and are not constant. The adjustable parameters include the heat transfer coefficient between the fuel and the coolant. The online calibration module receives the parameters output by the data acquisition module and the adjustable parameters determined by the adjustable parameter reconstruction module. Based on the model predictive control concept and intelligent optimization algorithm, it adjusts the adjustable parameters to realize the online calibration of the digital twin model and realize the mapping of the key parameters of the real reactor. The online correction module uses the core inlet working fluid temperature and control rod position as input values ​​for the digital twin model, and the measured core power as the tracking target value for model predictive control. It reconstructs adjustable parameters through intelligent optimization algorithms and performs online correction on the power and core outlet working fluid temperature solved by the digital twin model, thereby mapping the key parameters of the real reactor.

[0008] By adopting the above technical solution, the present invention has the following beneficial effects compared with the prior art. Of course, any product implementing the present invention does not necessarily need to achieve all of the following advantages at the same time: This invention provides an online calibration method for nuclear reactors based on a model predictive control architecture. The method uses sensors placed at various locations in the reactor to obtain the real-time state of the reactor as the target value for model predictive control tracking. By optimizing the algorithm, the values ​​of adjustable parameters are reconstructed, thereby performing online calibration of the power and core outlet working fluid temperature obtained by the digital twin model, ensuring that the parameters of the reactor are mapped under various operating conditions.

[0009] The specific embodiments of the present invention will now be described in further detail with reference to the accompanying drawings. Attached Figure Description

[0010] The accompanying drawings described below are merely some embodiments. Those skilled in the art can obtain other drawings based on these drawings without any creative effort. In the drawings: Figure 1 This is a flowchart of an online calibration method for nuclear reactors based on a model predictive control architecture, according to the present invention. Figure 2 This is a flowchart of the S101 method in this invention; Figure 3 This is a flowchart of the S201 method in this invention; Figure 4 This is a flowchart of the S301 method in this invention; Figure 5 This is a flowchart of the S401 method in this invention; Figure 6 This is a block diagram of an online calibration system for nuclear reactors based on a model predictive control architecture, according to the present invention. Figure 7 This is a diagram showing the rolling optimization structure of the algorithm used for online correction in this invention; Figure 8 This study compares the key response parameters of an integrated pressurized water reactor under different operating conditions with those of a real reactor after using the control system of this invention.

[0011] It should be noted that these accompanying drawings and textual descriptions are not intended to limit the scope of the invention in any way, but rather to illustrate the concept of the invention to those skilled in the art by referring to specific embodiments. Detailed Implementation

[0012] The invention will now be described in further detail with reference to the accompanying drawings.

[0013] Please see Figure 1-8 As shown, this embodiment provides an online calibration method for nuclear reactors based on a model predictive control architecture, including the following steps: S101. Various sensors, such as neutron flux, temperature, and control rod position, are arranged in the reactor system. S201. Obtain the sensor measurements from S101, such as core power, core inlet working fluid temperature, and control rod position. S301. Establish a digital twin model of the reactor; S401. Substitute the measured values ​​from S201 into the mathematical model in S301 to obtain the output of the digital twin model calculation. Adjust the adjustable parameters of the digital twin model to achieve online correction.

[0014] In one specific embodiment, refer to Figure 2 As shown, S101 includes the following steps: S1011. Arrange neutron flux measurement sensors inside the reactor pressure vessel to measure the neutron flux distribution in the reactor core. S1012. Temperature sensors are installed on the surface of the reactor outlet nozzle to measure the temperature of the working fluid at the reactor core outlet.

[0015] S1013. Sensors are arranged in the reactor control rod drive mechanism to measure the position of the control rods.

[0016] In one specific embodiment, refer to Figure 3 As shown, S201 includes the following steps: S2011. Obtain the neutron flux values ​​at various points in the core measured by the neutron flux measurement sensor, and calculate the core power value accordingly. S2012. Obtain the core inlet working fluid temperature measured by the temperature sensor.

[0017] S2013. Obtain the position of the control rod as measured by the control rod sensor.

[0018] In one specific embodiment, refer to Figure 4 As shown, S301 includes the following steps: S3011. Construct a reactor core point dynamics model; S3012. Construct a reactor thermal-hydraulic model, including fuel temperature and coolant temperature; S3013. Evaluate the parameters of the established digital twin model of the nuclear reactor, preliminarily screen adjustable parameters, conduct sensitivity analysis, and finally determine the adjustable parameters for online calibration.

[0019] In one specific embodiment, refer to Figure 5 As shown, S401 includes the following steps: S4011. Input the control rod position measured by the control rod sensor in S2013 into the reactor core point dynamics model in S3011 to obtain the power calculation value of the digital twin model. S4012. Input the core inlet working fluid temperature value measured by the temperature sensor in S2012 into the reactor thermal-hydraulic model in S3012 to obtain the calculated core outlet working fluid temperature value of the digital twin model. S4013. The power and core outlet working fluid temperature calculated by the digital twin model are compared with the actual measured parameters of the real reactor. The calculation results of the digital twin model are adjusted by combining the idea of ​​model predictive control with some optimization algorithms, thereby achieving online correction and mapping the reactor state.

[0020] In one specific embodiment, the specific content of S4013 is as follows: This paper utilizes the principles of Model Predictive Control (MPC) and employs Particle Swarm Optimization (PSO) or other optimization algorithms to optimize adjustable parameters, ultimately forming an MPC-PSO rolling optimization architecture. The core idea of ​​this optimization algorithm is to randomly initialize the positions and velocities of the particle swarm, with each particle representing a set of solutions for the adjustable parameters. The optimal solution is then sought through subsequent iterations. The velocity update formula for the i-th particle during the k-th iteration is shown below:

[0021]

[0022] In the formula, gbest represents the globally optimal particle, pbest represents the optimal particle in each generation, v is the particle update velocity vector, w is the inertia weight, c1 and c2 are learning factors, and rand(0,1) represents a random number between 0 and 1. Reference Figure 6 As shown, the present invention also discloses a nuclear reactor online calibration method based on a model predictive control architecture. The method includes a data acquisition module, a reactor adjustable parameter reconfiguration module, and an online calibration module connected in sequence. The data acquisition module is used to acquire the measurement values ​​of sensors located at various locations in the reactor. The reactor adjustable parameter reconfiguration module is used to screen adjustable parameters in the digital twin model and perform sensitivity analysis to prepare for online calibration. The online calibration module acquires sensor measurements and reconstructed values ​​of adjustable parameters, and then performs online calibration through model predictive control and optimization algorithms to achieve mapping of key parameters of a real reactor.

[0023] In one specific embodiment, the sensor measurements acquired by the data acquisition module include: core power measured by the neutron flux sensor, core outlet working fluid temperature measured by the temperature sensor, and control rod position measured by the control rod sensor.

[0024] In one specific embodiment, the online calibration module uses the core working fluid inlet temperature measured by the temperature sensor and the control rod position measured by the control rod sensor as input values ​​for the digital twin model, and the core power measured by the sensor as the target value for model predictive control tracking. The values ​​of adjustable parameters are reconstructed through optimization algorithms, thereby performing online calibration on the power and core outlet working fluid temperature solved by the digital twin model, and realizing the mapping of key parameters of the real reactor.

[0025] The rolling optimization structure of the algorithm used in the online correction in this invention is as follows: Figure 7 As shown; Model rolling calibration method such as Figure 8 As shown, a reconstructable digital twin model is used as the predictive model to estimate the output response for specific control variables. A solution algorithm (such as PSO) is then applied to determine a set of control variables to minimize the computational bias of the model; these control variables are subsequently used to adjust the target model.

[0026] Figure 8 -(a) shows the nuclear power comparison curves of the integrated pressurized water reactor under the 100%-20%FP step load reduction condition; Figure 8 -(b) represents the relative error of online nuclear power correction for the integrated pressurized water reactor under the 100%-20%FP step load reduction condition; Figure 8-(c) shows the coolant outlet temperature comparison curves of the integrated pressurized water reactor under the 100%-20%FP step load reduction condition; Figure 8 -(d) represents the online correction relative error of coolant outlet temperature of the integrated pressurized water reactor under the 100%-20%FP step load reduction condition; Figure 8 -(e) shows the fuel temperature comparison curves of the integrated pressurized water reactor under the 100%-20%FP step-down load condition; Figure 8 -(f) represents the relative error of online fuel temperature correction for the integrated pressurized water reactor under the 100%-20%FP step load reduction condition; The results above demonstrate that this method, which combines particle swarm optimization with model predictive control, can effectively create a digital twin of a real reactor by using the actual reactor state as the target. This ensures low relative error, enhances reactor safety and reliability, and provides a new solution for the design of digital twin systems for real reactors.

[0027] It should be noted that all electrical devices involved in this application can be powered by batteries or external power sources. This invention is not limited to the embodiments described above. Anyone should understand that structural changes made under the guidance of this invention, and any technical solutions that are the same as or similar to this invention, fall within the protection scope of this invention. Technical aspects, shapes, and structures not described in detail in this invention are all publicly known technologies.

Claims

1. A method for online calibration of nuclear reactors based on a model predictive control architecture, characterized in that, The adjustable parameters of the digital twin model are optimized using model predictive control and intelligent optimization algorithms to achieve reactor state mapping. This includes the following steps: S101. Neutron flux sensors, temperature sensors and control rod position sensors are installed at designated locations in the reactor system. S201. Collect the measurement data from the sensor and process it to obtain the core power, core inlet working fluid temperature and control rod position parameters; S301. Construct a digital twin model of the reactor, and screen and determine the adjustable parameters of the model; S401. Input the parameters into the digital twin model to obtain the calculated output value, and adjust the adjustable parameters through model predictive control and intelligent optimization algorithms to complete the online correction.

2. The online calibration method for nuclear reactors based on a model predictive control architecture according to claim 1, characterized in that, S101 specifically includes: S1011. A neutron flux measurement sensor is installed inside the reactor pressure vessel to measure the neutron flux distribution in the reactor core in real time. S1012. A temperature sensor is installed on the surface of the reactor outlet nozzle to measure the temperature of the working fluid at the reactor core outlet in real time. S1013. Displacement sensors are arranged on the reactor control rod drive mechanism to measure the position of the control rods in real time.

3. The online calibration method for nuclear reactors based on a model predictive control architecture according to claim 1, characterized in that, S201 specifically includes: S2011. Obtain neutron flux data from various locations in the reactor core collected by the neutron flux measurement sensor, and calculate the reactor core power value accordingly. S2012. Obtain core inlet working fluid temperature data directly measured by temperature sensors; S2013. Obtain the control rod position data measured by the control rod position sensor.

4. The online calibration method for nuclear reactors based on a model predictive control architecture according to claim 1, characterized in that, S301 specifically includes: S3011. Construct a reactor core point dynamics model; S3012. Construct a reactor thermal-hydraulic model that includes a fuel temperature calculation module and a coolant temperature calculation module; S3013. Parameter evaluation is performed on the constructed digital twin model of the nuclear reactor, and adjustable parameters are initially screened. The final adjustable parameters for online calibration are determined through sensitivity analysis.

5. The online calibration method for nuclear reactors based on a model predictive control architecture according to claim 1, characterized in that, Specifically, S401 includes: S4011. Input the control rod position data obtained in S2013 into the reactor core point dynamics model constructed in S3011 to calculate the core power calculation value of the digital twin model. S4012. Input the core inlet working fluid temperature data obtained in S2012 into the reactor thermal-hydraulic model constructed in S3012 to calculate the core outlet working fluid temperature of the digital twin model. S4013. The calculated values ​​of core power and core outlet working fluid temperature are compared with the measured core power value in S2011 and the measured core outlet working fluid temperature value in S1012, respectively. Based on the model predictive control concept and combined with intelligent optimization algorithm, the calculation results of the digital twin model are adjusted to achieve online correction and map the reactor state.

6. A nuclear reactor online calibration system based on a model predictive control architecture, characterized in that, The online calibration method according to any one of claims 1-5 includes a data acquisition module, a reactor adjustable parameter reconfiguration module, and an online calibration module connected in sequence: The data acquisition module is used to acquire measurement data from sensors located throughout the reactor. The reactor adjustable parameter reconfiguration module is used to screen adjustable parameters in the digital twin model, determine the final adjustable parameters through sensitivity analysis, and provide a parameter basis for online calibration. The online calibration module receives parameters output by the data acquisition module and adjustable parameters determined by the adjustable parameter reconstruction module. Based on the concept of model predictive control and intelligent optimization algorithms, it adjusts the adjustable parameters to achieve online calibration of the digital twin model and realize the mapping of key parameters of the real reactor.

7. The online calibration system for nuclear reactors based on a model predictive control architecture according to claim 6, characterized in that, The sensor measurements acquired by the data acquisition module include: core power measured by the neutron flux sensor, core outlet working fluid temperature measured by the temperature sensor, and control rod position measured by the control rod sensor.

8. The online calibration system for nuclear reactors based on a model predictive control architecture according to claim 6, characterized in that, The reactor adjustable parameter reconfiguration module is used to screen adjustable parameters in the digital twin model. These adjustable parameters are affected by the core power and are not constant.

9. The online calibration system for a nuclear reactor based on a model predictive control architecture according to claim 8, characterized in that, The adjustable parameters include the heat transfer coefficient between the fuel and the coolant.

10. The online calibration system for a nuclear reactor based on a model predictive control architecture according to claim 6, characterized in that, The online correction module uses the core inlet working fluid temperature and control rod position as input values ​​for the digital twin model, and the measured core power as the tracking target value for model predictive control. It reconstructs adjustable parameters through intelligent optimization algorithms and performs online correction on the power and core outlet working fluid temperature solved by the digital twin model, thereby achieving mapping of the key parameters of the real reactor.

Images