An interactive multi-model weighted fusion method for core power control

Abstract

This invention belongs to the technical field of lead-bismuth cooled reactor control system design, and specifically relates to an interactive multi-model weighted fusion method for core power control. The invention includes the following steps: S1, constructing a core model and performing lead-bismuth heat transfer calculations; S2, constructing transfer functions, building four power step transfer functions to characterize the lead-bismuth reactor model, converting the transfer functions into state-space form to obtain a linearized model; S3, dynamically mixing the initial states of each model using Markov transition probabilities, independently calculating the predicted values ​​and residuals of each model using Kalman filtering; updating the model weights in real time based on residual covariance and Bayes' theorem, and weighting and fusing the outputs of each model. This invention obtains the system's transfer function from input to output based on the constructed core model, fuses real-time measurement data through a Markov transition mechanism, and weights and fuses the outputs of each model, ultimately achieving real-time matching of the optimal model to support core power control in lead-bismuth reactors.

Description

Technical Field

[0001] This invention belongs to the field of lead-bismuth cooled reactor control system design technology, and relates to an interactive multi-model weighted fusion method for core power control. Background Technology

[0002] The density of lead-bismuth coolant can vary by up to 20% within the temperature range of 300℃-800℃. This leads to severe temperature feedback nonlinearity in lead-bismuth reactors during operation, resulting in linearization model errors exceeding 30% when switching between power steps of 100%FP / 75%FP / 50%FP / 25%FP (%FP: rated power percentage, abbreviated as %). The strong nonlinearity and multi-condition switching challenges posed by this process pose significant challenges to the control of lead-bismuth reactors. A multi-model approach, using weighted fusion of measured data, can be employed to achieve optimal model matching.

[0003] 1. Existing technology

[0004] Reference [1] "An Advanced Control System and Design Method for a Lead-Bismuth Stack" (Invention Patent, Wei Xinyu, Zhang Ru, et al., Application Date: 2024-12-10). This reference mainly relates to an advanced control system and design method for a lead-bismuth stack. The method includes building a lead-bismuth stack model, which includes two controlled object models: the stack core and the steam generator. Based on the built lead-bismuth stack model, the characteristics of the controller output anti-disturbance and anti-noise capability are analyzed for various advanced control algorithms, and the characteristic analysis results corresponding to each advanced control algorithm are obtained. Based on the characteristic analysis results, the superior advanced control algorithm is selected. Based on the superior advanced control algorithm, the stack core power self-disturbance controller, the steam pressure internal model controller and the main steam flow internal model controller are designed, thereby designing the advanced control system for the lead-bismuth stack. The advanced control system designed by this method not only has good response speed and stability, but also ensures that the lead-bismuth stack can meet the application requirements of complex operating environment and changing working conditions.

[0005] Reference [2] "Research on Simulation Model of Small Integrated Lead-Bismuth Cooled Reactor" (Academic Paper, Sun Yuanli, Song Zhihao and Lü Xiangbo et al., Received: 2022-12-09). This paper takes the integrated small lead-bismuth cooled reactor as the research object, establishes a steam generator secondary side model, a primary loop main cooling system model, a constitutive model and a proportional-integral-derivative (PID) control model based on the four-equation drift flow model, and conducts research on the operation control characteristics of the lead-bismuth cooled reactor. The results show that the steady-state calculation results are in good agreement with the design values, and the model can accurately simulate the operation characteristics of the lead-bismuth cooled reactor; the system parameter overshoot is small under rapid load change operation conditions, and the reactor power can follow the rapid change of steam flow; the blockage of heat transfer tubes has a significant impact on reactor operation, and the steam flow rate decreases by about 6.7% for each blocked heat transfer tube.

[0006] 2. Limitations of the aforementioned background technology

[0007] References [1][2] both involve the construction of lead-bismuth cooled reactor models. The control object models involve the reactor core and steam generator. They do not address the control issues such as strong nonlinearity and multi-condition switching of reactor core power during multi-power step switching. They all use a set of system parameters to simulate and characterize the reactor core. However, the system parameters differ greatly between the low-power step and the high-power step. The relevant modeling methods in the background technology have high model errors and large nonlinearity, which are not conducive to the design and simulation verification of the control system. Summary of the Invention

[0008] The technical problem solved by this invention is to provide an interactive multi-model weighted fusion method for core power control, which addresses the challenges of strong nonlinearity and multi-condition switching in the collaborative control of lead-bismuth reactor core power control. Based on the constructed core model, the transfer function from input to output of the system is obtained through linearization and Laplace transform. Step transfer functions are applied for four basic power levels: 100%, 75%, 50%, and 25%. The outputs of each model are weighted and fused by real-time measurement data through a Markov transfer mechanism, ultimately achieving real-time matching of the optimal model to support the core power control of lead-bismuth reactors.

[0009] The technical solution adopted in this invention is as follows:

[0010] A method for controlling the core power and average coolant temperature of a lead-bismuth cooled reactor includes the following steps:

[0011] S1. Construct a core model and perform lead-bismuth heat transfer calculations;

[0012] S2. Construct transfer functions. Based on the actual operation of the lead-bismuth pile, four basic power operation steps of 100%, 75%, 50%, and 25% are obtained. Four power step transfer functions are constructed to characterize the lead-bismuth pile model. The transfer functions are transformed into state-space form to obtain a linearized model.

[0013] S3. Parallel operation of linearized models of lead-bismuth piles at four basic power operation stages of 100%, 75%, 50%, and 25% is performed. The initial states of each model are dynamically mixed using Markov transition probabilities, and the predicted values ​​and residuals of each model are calculated independently using Kalman filtering. Subsequently, the model weights are updated in real time based on residual covariance and Bayes' theorem. Finally, the outputs of each model are weighted and fused to achieve real-time matching of the best model.

[0014] The core model includes a point reactor dynamics model, a reactive input model, and a thermodynamic dynamics model.

[0015] Constructing a point heap dynamics model:

[0016] The point-pile dynamics model includes six sets of delayed neutrons, and the point-pile equations are as follows:

[0017]

[0018]

[0019] In the formula, n(t) represents the neutron density / neutron number in m. -3 ρ(t) - Total reactivity within the reactor core / dk·k -1 C i (t) - Concentration of precursor nuclei of the i-th group of delayed neutrons / m -3 ;λ i - Decay constant of the i-th precursor nucleus / s; β i - The share of delayed neutron emission in group i, %; Λ - Neutron generation time / s.

[0020] Constructing a reactive computational input model:

[0021] The reactivity calculation input model includes the following reactivity: reactivity introduced by control rod movement, fuel temperature effect (Doppler effect), coolant temperature effect, reactivity caused by axial geometry changes in the core, reactivity caused by radial geometry changes in the core, burnup reactivity, and other reactivity. The specific models are as follows:

[0022] ρ=ρ Rod +ρ f,T +ρ c,T +ρ c,d +ρ b +ρ a +ρ r +ρ Else

[0023] In the formula, ρ Rod -Reactivity caused by the movement of the control rod / pcm; ρ f,T-Reactivity caused by fuel temperature changes / pcm; ρ c,T -Reactivity caused by coolant temperature change / pcm; ρ c,d -Reactivity caused by changes in coolant density / pcm; ρ b -Reactivity caused by fuel burnup effect / pcm; ρ a -Reactivity caused by axial geometric changes in the reactor core / pcm; ρ r -Reactivity caused by radial geometric changes in the reactor core / pcm; ρ Else - indicates other reactivity / pcm. This item is for the convenience of model correction and is set to 0 by default unless otherwise specified.

[0024] Constructing a thermodynamic model

[0025] The thermodynamic model includes a thermohydraulic model and a core heat transfer model. The thermohydraulic model adopts a single-channel lumped parameter model, and the core heat transfer model adopts the Mann model.

[0026] According to the law of conservation of energy, the dynamic heat transfer equation between fuel and coolant is:

[0027]

[0028] In the formula, T f - Average fuel temperature / °C; T m - Average temperature of the coolant / °C; T in - Coolant inlet temperature / °C; T out - Coolant outlet temperature / °C; f f - Fuel thermal coefficient; P0 - Rated power / W; μ f - Heat capacity of fuel / J·℃ -1 μ f =m f C p,f μ c -Heat capacity of coolant / J·℃ -1 μ c =m c C p,c C p,f and C p,c -Specific heat capacity at constant pressure of fuel and coolant / J·kg -1 ·℃ -1 Ω - Heat transfer coefficient between fuel and coolant / W·℃ -1 ;τ c -Delay time / s, τ c =μ c / (W p C p,c );W p- Coolant flow rate / kg·s -1 .

[0029] The lead-bismuth heat transfer calculation:

[0030] The simulation is simplified based on the following assumptions:

[0031] The coolant is fully mixed in the hot and cold lead-bismuth pools, meaning the outlet temperature is equal to the average temperature of the hot and cold pools; the hot and cold lead-bismuth pools are insulated from each other.

[0032] Based on the law of conservation of energy, the energy equation for hot and cold lead-bismuth pools is:

[0033]

[0034] In the formula: T hot (t), T cold (t) - Outlet temperature of hot and cold lead-bismuth baths / °C; T hot,in (t), T cold,in (t) - Inlet temperature of hot and cold lead-bismuth baths / °C; F hot F cold - Volumetric flow rate of hot and cold lead-bismuth baths / m³ 3 ·s -1 V hot V cold -Effective capacity of hot and cold lead-bismuth pools / m 3 .

[0035] In step 2, based on the core model constructed in S1, the inputs are selected as the core inlet coolant temperature, the reactivity introduced by the control rods, and the coolant mass flow rate, and the outputs are selected as the core outlet coolant temperature and normalized power. The transfer function of the system from input to output is obtained through linearization and Laplace transform, and the linearized model of the lead-bismuth reactor at four basic power operating steps of 100%, 75%, 50%, and 25% is obtained.

[0036] Linearized model of the 100% operational stage of the lead-bismuth pile:

[0037] Three-input two-output means there are 6 transfer functions at each power level, where T is the transfer function. in The value represents the core coolant inlet temperature, rod represents the reactivity introduced by the control rods, W represents the coolant mass flow rate, and T represents the temperature at the reactor core inlet. out Here, P represents the core outlet coolant temperature, and P represents the normalized power. The system transfer function at 100% power level is as follows:

[0038]

[0039]

[0040] S3 specifically includes:

[0041] Establish the transition probability matrix;

[0042] Initial model probability;

[0043] Calculate the matching degree of each model;

[0044] Model probability update;

[0045] The outputs of each model are weighted and fused.

[0046] The transition probability matrix:

[0047]

[0048] The matching degree of each model is calculated using log-likelihood calculation based on real-time measurement data to avoid numerical underflow.

[0049]

[0050] in:

[0051] Let y represent the measurement residual of the i-th model. k For real-time measurement data, These are the state estimation parameters;

[0052] Let H be the residual covariance matrix, H be the observation matrix, P be the prediction covariance, and R be the measurement noise covariance.

[0053] The model probability update:

[0054] The Markov transition mechanism is used to achieve smooth switching between models and update the model probabilities through probability transition matrix and dynamic weight updates.

[0055] Markov transition mechanism:

[0056]

[0057] A probability correction mechanism was developed based on the actual operation of the lead-bismuth pile.

[0058] Power change rate constraint: When dP / dt > 5% / min | dP / dt | > 5% / min, the freeze probability is updated;

[0059] Minimum probability threshold: u i When the value is less than 0.05, it is forcibly set to zero to avoid numerical instability;

[0060] Furthermore, a coolant density gradient correction factor is introduced:

[0061] π ij ←π ij×exp(-|Δρ / ρ0|)

[0062] Δρ is the rate of change of density;

[0063] The outputs of the weighted fusion models:

[0064] Mixed state estimation:

[0065]

[0066] Covariance fusion:

[0067]

[0068] The beneficial effects of this invention are:

[0069] (1) The present invention provides an interactive multi-model weighted fusion method for core power control. In response to the strong nonlinearity problem caused by the large change in coolant density when switching power steps in lead-bismuth reactors, the method establishes local linearization models for four power steps of 100%, 75%, 50%, and 25%, which avoids the large error of a single model under high / low power conditions and significantly improves the model accuracy.

[0070] (2) The present invention provides an interactive multi-model weighted fusion method for core power control. Through parallel calculation and dynamic fusion of multiple models, and by introducing a coolant density gradient correction factor and a power change rate constraint for probabilistic correction, it achieves accurate prediction of core outlet coolant temperature and normalized power, provides real-time adaptive model parameters for the control system, and improves the control performance and stability of lead-bismuth reactors under complex operating conditions. Attached Figure Description

[0071] To more clearly illustrate the embodiments of the present invention, the accompanying drawings used in describing the embodiments of the present invention will be briefly described below. Obviously, the drawings described below are merely some embodiments recorded in the present invention. Those skilled in the art can derive other drawings from the following drawings without any creative effort.

[0072] Figure 1 A flowchart of an interactive multi-model weighted fusion method for core power control. Detailed Implementation

[0073] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of the present invention.

[0074] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., refer to the orientation or positional relationship shown in the accompanying drawings, and are used only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0075] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or a connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0076] like Figure 1 As shown, the present invention provides a method for controlling the core power and average coolant temperature of a lead-bismuth cooled reactor, comprising the following steps:

[0077] S1. Construct a core model and perform lead-bismuth heat transfer calculations.

[0078] The core model includes a point reactor dynamics model, a reactive input model, and a thermodynamic dynamics model;

[0079] S101. Constructing a point heap dynamics model

[0080] The point-pile dynamics model includes six sets of delayed neutrons, and the point-pile equations are as follows:

[0081]

[0082] In the formula, n(t) represents the neutron density / neutron number in m. -3 ρ(t) - Total reactivity within the reactor core / dk·k -1 C i (t) - Concentration of precursor nuclei of the i-th group of delayed neutrons / m -3 ;λ i - Decay constant of the i-th precursor nucleus / s; β i - The share of delayed neutron emission in group i, %; Λ - Neutron generation time / s.

[0083] S102. Constructing a reactive computational input model

[0084] The reactivity within the reactor core plays a crucial role in reactor startup, operation, and regulation. The reactivity calculation input model constructed in this embodiment includes the following reactivity factors: reactivity introduced by control rod movement, fuel temperature effect (Doppler effect), coolant temperature effect, reactivity caused by axial geometric changes in the core, reactivity caused by radial geometric changes in the core, burnup reactivity, and other reactivity factors. The specific model is as follows:

[0085] ρ=ρ Rod +ρ f,T +ρ c,T +ρ c,d +ρ b +ρ a +ρ r +ρ Else

[0086] In the formula, ρ Rod -Reactivity caused by the movement of the control rod / pcm; ρ f,T -Reactivity caused by fuel temperature changes / pcm; ρ c,T -Reactivity caused by coolant temperature change / pcm; ρ c,d -Reactivity caused by changes in coolant density / pcm; ρ b -Reactivity caused by fuel burnup effect / pcm; ρ a -Reactivity caused by axial geometric changes in the reactor core / pcm; ρ r -Reactivity caused by radial geometric changes in the reactor core / pcm; ρ Else - indicates other reactivity / pcm. This item is for the convenience of model correction and is set to 0 by default unless otherwise specified.

[0087] S103, Constructing a thermodynamic model

[0088] The thermodynamic model includes a thermohydraulic model and a core heat transfer model. The thermohydraulic model adopts a single-channel lumped parameter model, and the core heat transfer model adopts the Mann model. This model uses one fuel node to correspond to two coolant nodes, which has higher accuracy than the traditional single-node method.

[0089] According to the law of conservation of energy, the dynamic heat transfer equation between fuel and coolant is:

[0090]

[0091]

[0092] In the formula, T f - Average fuel temperature / °C; T m - Average temperature of the coolant / °C; T in- Coolant inlet temperature / °C; T out - Coolant outlet temperature / °C; f f - Fuel thermal coefficient; P0 - Rated power / W; μ f - Heat capacity of fuel / J·℃ -1 μ f =m f C p,f μ c -Heat capacity of coolant / J·℃ -1 μ c =m c C p,c C p,f and C p,c -Specific heat capacity at constant pressure of fuel and coolant / J·kg -1 ·℃ -1 Ω - Heat transfer coefficient between fuel and coolant / W·℃ -1 ;τ c -Delay time / s, τ c =μ c / (W p C p,c );W p - Coolant flow rate / kg·s -1 .

[0093] S104, Lead-bismuth heat transfer calculation

[0094] The lead-bismuth heat exchange process is as follows: after the coolant flows out of the reactor core, it exchanges heat with the steam generator through a hot lead-bismuth pool. The cooled lead-bismuth then flows back into the reactor core through a cold lead-bismuth pool to complete the cycle. To simplify the simulation, the following assumptions are made:

[0095] (1) The coolant is fully mixed in the hot and cold lead-bismuth pools, that is, the outlet temperature is equal to the average temperature of the hot and cold pools;

[0096] (2) The cold and hot lead-bismuth pools are insulated.

[0097] Based on the law of conservation of energy, the energy equation for hot and cold lead-bismuth pools can be obtained as follows:

[0098]

[0099] In the formula: T hot (t), T cold (t) - Outlet temperature of hot and cold lead-bismuth baths / °C; T hot,in (t), T cold,in (t) - Inlet temperature of hot and cold lead-bismuth baths / °C; F hot F cold - Volumetric flow rate of hot and cold lead-bismuth baths / m³ 3 ·s -1 V hotV cold -Effective capacity of hot and cold lead-bismuth pools / m 3 .

[0100] S2. Constructing transfer functions: Based on the actual operation of the lead-bismuth pile, four basic power operation steps (100%, 75%, 50%, and 25%) are obtained. Transfer functions for these four power steps are constructed to characterize the lead-bismuth pile model. The transfer functions are then transformed into a state-space form to obtain a linearized model.

[0101] Based on the core model constructed in S1, the inputs are selected as the core inlet coolant temperature, the reactivity introduced by the control rods, and the coolant mass flow rate. The outputs are selected as the core outlet coolant temperature and the normalized power. The transfer function of the system from input to output can be obtained through linearization and Laplace transform.

[0102] Three-input two-output means there are 6 transfer functions at each power level, where T is the transfer function. in The value represents the core coolant inlet temperature, rod represents the reactivity introduced by the control rods, W represents the coolant mass flow rate, and T represents the temperature at the reactor core inlet. out Here, P represents the core outlet coolant temperature, and P represents the normalized power. The system transfer function at 100% power level is as follows:

[0103]

[0104]

[0105] S3, Interactive Multi-Model Weighted Fusion

[0106] Linearized models of a lead-bismuth pile operating at four basic power levels (100%, 75%, 50%, and 25%) are run in parallel. The initial states of each model are dynamically mixed using Markov transition probabilities, and the predicted values ​​and residuals of each model are independently calculated using Kalman filtering. Subsequently, the model weights are updated in real-time based on residual covariance and Bayes' theorem. Finally, the outputs of each model are weighted and fused to achieve real-time matching of the optimal model.

[0107] S301 Establish the transition probability matrix

[0108] Based on the operating characteristics and power conversion characteristics of lead-bismuth stacks, and considering that the high-power switching probability is higher than the low-power range, the transition probability matrix is ​​designed as follows:

[0109]

[0110] S302 Initial Model Probability

[0111] The initial model probability refers to the initial weight or confidence level assigned to each candidate model at startup (k=0). It reflects the system's prior confidence in each model before any observation data is obtained.

[0112] Considering the initial state is more likely to be at low power:

[0113] μ0 = [0.1 0.2 0.3 0.4]

[0114] S303 calculates the matching degree (likelihood function) for each model.

[0115] Based on real-time measurement data (such as neutron flux and coolant temperature), the matching degree (likelihood function) of each model is calculated to quantify the degree of matching between the current observation data and each model.

[0116] Log-likelihood calculations are used to avoid numerical underflow.

[0117]

[0118] in:

[0119] Let y represent the measurement residual of the i-th model. k For real-time measurement data, These are the state estimation parameters;

[0120] Let H be the residual covariance matrix, H be the observation matrix, P be the prediction covariance, and R be the measurement noise covariance.

[0121] S304 Model Probability Update

[0122] The aforementioned method utilizes the Markov transition mechanism to achieve smooth switching between models and updates model probabilities through probability transition matrices and dynamic weight updates. Markov transition mechanism:

[0123]

[0124] A probability correction mechanism was developed based on the actual operation of the lead-bismuth pile.

[0125] Power change rate constraint: When dP / dt > 5% / min | dP / dt | > 5% / min, the freeze probability is updated;

[0126] Minimum probability threshold: u i When the value is less than 0.05, it is forcibly set to zero to avoid numerical instability;

[0127] Furthermore, a coolant density gradient correction factor is introduced:

[0128] π ij ←π ij ×exp(-|Δρ / ρ0|)

[0129] (Δρ is the rate of change of density)

[0130] S305 weighted fusion of the outputs of each model

[0131] Mixed state estimation:

[0132]

[0133] Covariance fusion:

[0134]

[0135] Smooth switching between models is achieved through a probability transition matrix and dynamic weight updates, updating the model probabilities. A probability correction mechanism is formulated based on the actual operation of the lead-bismuth reactor, introducing power variation constraints and coolant density gradient correction factors. Finally, the outputs of each model are weighted and fused to achieve real-time matching of the optimal model, supporting the core power control of the lead-bismuth reactor.

[0136] Twenty-four transfer functions for four power steps (100%, 75%, 50%, and 25%) are generated through linearization and Laplace transform. Markov transition probability matrices are then designed (e.g., a probability of 0.5 for 100% → 75%). Model weights are updated in real-time using Kalman filter residual covariance and Bayes' theorem, and probabilistic corrections are made using coolant density gradient correction factors and power change rate constraints. Finally, the optimal control quantity is output through the fusion of mixed state estimation and covariance, achieving adaptive switching across multiple operating conditions and significantly reducing model errors.

[0137] While those skilled in the art will recognize that the present invention is not limited to the details of the exemplary embodiments described above, and that it can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention, the embodiments should be considered illustrative and non-limiting in all respects. The scope of the invention is defined by the appended claims rather than the foregoing description, and therefore all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0138] Furthermore, it should be understood that although the present invention is described according to embodiments, not every embodiment contains only one independent technical solution. This way of describing the specification is only for clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A method for controlling the core power and average coolant temperature of a lead-bismuth cooled reactor, characterized in that, Includes the following steps: S1. Construct a core model and perform lead-bismuth heat transfer calculations; S2. Construct transfer functions. Based on the actual operation of the lead-bismuth pile, four basic power operation steps of 100%, 75%, 50%, and 25% are obtained. Four power step transfer functions are constructed to characterize the lead-bismuth pile model. The transfer functions are transformed into state-space form to obtain a linearized model. S3. Parallel operation of linearized models of lead-bismuth piles at four basic power operation stages of 100%, 75%, 50%, and 25% is performed. The initial states of each model are dynamically mixed using Markov transition probabilities, and the predicted values ​​and residuals of each model are calculated independently using Kalman filtering. Subsequently, the model weights are updated in real time based on residual covariance and Bayes' theorem. Finally, the outputs of each model are weighted and fused to achieve real-time matching of the best model.

2. The method for controlling the core power and average coolant temperature of a lead-bismuth cooled reactor according to claim 1, characterized in that, The core model includes a point reactor dynamics model, a reactive input model, and a thermodynamic dynamics model.

3. The method for controlling the core power and average coolant temperature of a lead-bismuth cooled reactor according to claim 2, characterized in that, Constructing a point heap dynamics model: The point-pile dynamics model includes six sets of delayed neutrons, and the point-pile equations are as follows: In the formula, n(t) represents the neutron density / neutron number in m. -3 ρ(t) - Total reactivity within the reactor core / dk·k -1 C i (t) - Concentration of precursor nuclei of the i-th group of delayed neutrons / m -3 ; λ i - Decay constant of the i-th precursor nucleus / s; β i - The share of delayed neutron emission in group i, %; Λ - Neutron generation time / s.

4. The method for controlling the core power and average coolant temperature of a lead-bismuth cooled reactor according to claim 3, characterized in that, Constructing a reactive computational input model: The reactivity calculation input model includes the following reactivity: reactivity introduced by control rod movement, fuel temperature effect (Doppler effect), coolant temperature effect, reactivity caused by axial geometry changes in the core, reactivity caused by radial geometry changes in the core, burnup reactivity, and other reactivity. The specific models are as follows: p=p Rod +r f,T +r c,T +r c,d +r b +r a +r r +r Else In the formula, ρ Rod -Reactivity caused by the movement of the control rod / pcm; ρ f,T -Reactivity caused by fuel temperature changes / pcm; ρ c,T -Reactivity caused by coolant temperature change / pcm; ρ c,d -Reactivity caused by changes in coolant density / pcm; ρ b -Reactivity caused by fuel burnup effect / pcm; ρ a -Reactivity caused by axial geometric changes in the reactor core / pcm; ρ r -Reactivity caused by radial geometric changes in the reactor core / pcm; ρ Else - indicates other reactivity / pcm. This item is for the convenience of model correction and is set to 0 by default unless otherwise specified.

5. The method for controlling the core power and average coolant temperature of a lead-bismuth cooled reactor according to claim 4, characterized in that, Constructing a thermodynamic model The thermodynamic model includes a thermohydraulic model and a core heat transfer model. The thermohydraulic model adopts a single-channel lumped parameter model, and the core heat transfer model adopts the Mann model. According to the law of conservation of energy, the dynamic heat transfer equation between fuel and coolant is: In the formula, T f - Average fuel temperature / °C; T m - Average temperature of the coolant / °C; T in - Coolant inlet temperature / °C; T out - Coolant outlet temperature / °C; f f - Fuel thermal coefficient; P0 - Rated power / W; μ f - Heat capacity of fuel / J·℃ -1 μ f =m f C p,f μ c -Heat capacity of coolant / J·℃ -1 μ c =m c C p,c C p,f and C p,c -Specific heat capacity at constant pressure of fuel and coolant / J·kg -1 ·℃ -1 Ω - Heat transfer coefficient between fuel and coolant / W·℃ -1 ;τ c -Delay time / s, τ c =μ c / (W p C p,c );W p - Coolant flow rate / kg·s -1 .

6. The method for controlling the core power and average coolant temperature of a lead-bismuth cooled reactor according to claim 5, characterized in that, The lead-bismuth heat transfer calculation: The simulation is simplified based on the following assumptions: The coolant is fully mixed in the hot and cold lead-bismuth pools, meaning the outlet temperature is equal to the average temperature of the hot and cold pools; the hot and cold lead-bismuth pools are insulated from each other. Based on the law of conservation of energy, the energy equation for hot and cold lead-bismuth pools is: In the formula: T hot (t), T cold (t) - Outlet temperature of hot and cold lead-bismuth baths / °C; T hot,in (t), T cold,in (t) - Inlet temperature of hot and cold lead-bismuth baths / °C; F hot F cold - Volumetric flow rate of hot and cold lead-bismuth baths / m³ 3 ·s -1 V hot V cold -Effective capacity of hot and cold lead-bismuth pools / m 3 .

7. The method for controlling the core power and average coolant temperature of a lead-bismuth cooled reactor according to claim 6, characterized in that, In step 2, based on the core model constructed in S1, the inputs are selected as the core inlet coolant temperature, the reactivity introduced by the control rods, and the coolant mass flow rate, and the outputs are selected as the core outlet coolant temperature and normalized power. The transfer function of the system from input to output is obtained through linearization and Laplace transform, and the linearized model of the lead-bismuth reactor at four basic power operating steps of 100%, 75%, 50%, and 25% is obtained.

8. The method for controlling the core power and average coolant temperature of a lead-bismuth cooled reactor according to claim 7, characterized in that, Linearized model of the 100% operational stage of the lead-bismuth pile: Three-input two-output means there are 6 transfer functions at each power level, where T is the transfer function. in The value represents the core coolant inlet temperature, rod represents the reactivity introduced by the control rods, W represents the coolant mass flow rate, and T represents the temperature at the reactor core inlet. out Here, P represents the core outlet coolant temperature, and P represents the normalized power. The system transfer function at 100% power level is as follows:

9. The method for controlling the core power and average coolant temperature of a lead-bismuth cooled reactor according to claim 8, characterized in that, S3 specifically includes: Establish the transition probability matrix; Initial model probability; Calculate the matching degree of each model; Model probability update; The outputs of each model are weighted and fused.

10. The method for controlling the core power and average coolant temperature of a lead-bismuth cooled reactor according to claim 9, characterized in that, The transition probability matrix: The matching degree of each model is calculated using log-likelihood calculation based on real-time measurement data to avoid numerical underflow. in: Let y represent the measurement residual of the i-th model. k For real-time measurement data, These are the state estimation parameters; Let H be the residual covariance matrix, H be the observation matrix, P be the prediction covariance, and R be the measurement noise covariance. The model probability update: The Markov transition mechanism is used to achieve smooth switching between models and update the model probabilities through probability transition matrix and dynamic weight updates. Markov transition mechanism: A probability correction mechanism was developed based on the actual operation of the lead-bismuth pile. Power change rate constraint: When dP / dt > 5% / min | dP / dt | > 5% / min, the freeze probability is updated; Minimum probability threshold: u i When the value is less than 0.05, it is forcibly set to zero to avoid numerical instability; Furthermore, a coolant density gradient correction factor is introduced: p ij ←p ij ×exp(-|Δρ / ρ0|) Δρ is the rate of change of density; The outputs of the weighted fusion models: Mixed state estimation: Covariance fusion:

Images