Combined type preheating system of fused salt storage tank

Abstract

The invention discloses a composite preheating system of a molten salt storage tank, which relates to the field of molten salt storage tank preheating, and comprises at least two heterogeneous heat source modules with different thermodynamic characteristics and response rates; the dynamic heat management unit is configured to receive the output of the heterogeneous heat source module and execute a self-adaptive preheating strategy based on molten salt storage tank structure stress field online inversion and heat shock risk estimation; the distributed preheating actuator array is controlled by the dynamic thermal management unit and performs multivariable coupling control based on the real-time temperature field and stress field distribution of the fused salt storage tank; wherein the optimization objectives of the self-adaptive preheating strategy at least simultaneously comprise the shortest preheating time, the minimum thermal stress and the minimum operation cost, and the dynamic thermal management unit solves the multi-objective optimization problem. And dynamically distributing the contribution degree of each heterogeneous heat source module and the action mode of the distributed preheating actuator array in real time.

Description

Technical Field

[0001] This invention relates to the field of molten salt storage tank preheating, specifically a composite preheating system for molten salt storage tanks. Background Technology

[0002] Molten salt storage tanks are key thermal storage devices in systems such as concentrated solar power (CSP) and industrial waste heat utilization. The preheating process before startup is crucial for ensuring safe system operation and extending equipment life. Insufficient or improper preheating can lead to problems such as molten salt solidification and blockage, excessive thermal stress on the tank structure, and even equipment damage. Therefore, the design and control strategies of the preheating system are core aspects of engineering applications.

[0003] Currently, existing technologies for preheating molten salt storage tanks mainly employ single heat sources or simple combinations of heat sources. For example, using a single electric heating system for preheating, while simple to control, results in high operating costs; or using traditional methods such as steam tracing or heat transfer oil circulation, which suffer from slow thermal response and poor temperature uniformity. In addition, some systems have attempted to introduce industrial waste heat as a preheating source to reduce energy costs, but this is usually only used as an auxiliary heat source, lacking a synergistic optimization mechanism with the main heat source, and often employing sequential switching or fixed-ratio heat management methods.

[0004] However, the aforementioned existing technologies have significant limitations: First, they lack real-time monitoring and active control of the thermal stress on the tank structure during preheating, making it difficult to avoid the risk of material fatigue and damage caused by local overheating or thermal shock; second, they fail to achieve dynamic optimal matching between multiple heat sources based on system state and external conditions, making it impossible to balance preheating efficiency with operational economy and equipment safety; finally, existing preheating systems generally lack self-learning and continuous optimization capabilities, with fixed control strategies that are difficult to adapt to complex and changing actual working conditions. Therefore, developing a composite preheating system capable of multi-objective collaborative optimization and possessing intelligent decision-making and adaptive capabilities is of significant engineering importance. Summary of the Invention

[0005] Based on this, the purpose of the present invention is to provide a composite preheating system for molten salt storage tanks to solve the technical problems of low preheating efficiency, high operating costs, high risk of equipment thermal stress, and poor system adaptability caused by the lack of multi-objective collaborative optimization in the prior art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: including at least two heterogeneous heat source modules with different thermodynamic properties and response rates; A dynamic thermal management unit is configured to receive the output of the heterogeneous heat source module and execute an adaptive preheating strategy based on online inversion of the stress field of the molten salt tank structure and prediction of thermal shock risk. And a distributed preheating actuator array, which is controlled by the dynamic thermal management unit and performs multivariable coupled control based on the real-time temperature field and stress field distribution of the molten salt tank; The optimization objectives of the adaptive preheating strategy include at least the shortest preheating time, the minimum thermal stress, and the lowest operating cost. The dynamic thermal management unit solves this multi-objective optimization problem to dynamically allocate the contribution of each heterogeneous heat source module and the operating mode of the distributed preheating actuator array in real time.

[0007] In summary, the present invention has the following main beneficial effects: This invention introduces the collaborative control of heterogeneous heat source modules and dynamic thermal management units to construct an adaptive preheating strategy based on digital twins and multi-objective optimization. This effectively overcomes the technical defects of existing technologies, such as insufficient thermal stress control, low efficiency of multi-heat source collaboration, and lack of self-optimization capabilities. It achieves multi-objective collaborative optimization to improve preheating efficiency and reduce operating costs under strict thermal stress constraints. At the same time, through continuous self-learning in the preheating process, the system has the ability to intelligently regulate and control complex operating conditions. Attached Figure Description

[0008] Figure 1 This is the control flowchart of the present invention; Figure 2 This is a system architecture diagram of the present invention; Figure 3 This is a diagram showing the preheating timing stages of the present invention; Figure 4 This is the multi-objective optimization diagram of the present invention; Figure 5 This is a diagram of the self-learning module of the present invention. Detailed Implementation

[0009] The following will be based on embodiments of the present invention. Figures 1-5 The technical solutions in the embodiments of the present invention will be clearly and completely described below. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0010] The embodiments of the present invention will now be described.

[0011] A composite preheating system for molten salt storage tanks, such as Figures 1-5 As shown, it includes: System overall architecture.

[0012] This system mainly comprises three functional units: a heterogeneous heat source module, a dynamic thermal management unit, and a distributed preheating actuator array. These three units are tightly coupled with the control system via a data bus, forming a closed-loop intelligent preheating system.

[0013] 2. Detailed description of the heterogeneous heat source module.

[0014] This module is designed to provide two or more heat sources that are complementary in thermodynamic properties and response rates.

[0015] Transient electrical heat source: Specifically, a high-power resistance heater or an electromagnetic induction heating device can be used. To achieve the performance described in claim 2, for example, a power density of 60 kW / m³ can be selected. 3 The silicon molybdenum rod electric heating element receives control signals through a solid-state relay and can achieve full power output from zero within 5 seconds, thus providing rapid and precise replenishment of heat.

[0016] Industrial process waste heat source: This heat source is connected to, for example, a flue gas waste heat boiler in a chemical plant or a cooling water system in a metallurgical plant. To stabilize its output, a buffer tank and temperature control system are installed at the inlet pipeline to ensure that the temperature fluctuation of the hot fluid supplied to the preheating system is controlled within ±5% of the rated value (e.g., 300°C), or within ±15°C. This high thermal inertia heat source provides stable but poorly adjustable basic heat.

[0017] The cascade heating subsystem (corresponding to claim 10) is key to the deep recovery of low-temperature waste heat. In specific implementation, it is integrated as an independent heat pump loop at the output end of the industrial process waste heat source. The system uses a non-azeotropic mixture of R1234ze(E) and R245fa in a specific ratio as the intermediate heat transfer medium. Through a working fluid component concentration adjustment device (such as an adjustable storage tank) linked to the control system, the evaporation temperature curve of the working fluid in the evaporator can be changed, allowing it to absorb low-temperature waste heat more smoothly (e.g., flue gas decreasing from 150°C to 80°C). At the same time, by adjusting the condensation pressure, the heat released during the condensation process can better match the preheating demand curve of the molten salt storage tank at low temperatures (e.g., from ambient temperature to 150°C), achieving nonlinear matching of the heat transfer process and significantly improving waste heat utilization efficiency.

[0018] 3. Detailed description of the dynamic thermal management unit.

[0019] This unit is the brain of the system, responsible for complex decision-making and optimization.

[0020] Digital Twin Module: During system initialization, this module establishes a high-fidelity finite element analysis virtual tank model based on the CAD model of the molten salt tank, material properties (such as the thermal conductivity, specific heat capacity, and elastic modulus of carbon steel Q345R), and boundary conditions. During preheating, temperature data is collected in real time through a distributed fiber optic sensor network (e.g., sensing fibers using OFDR technology, with a measuring point every 10 cm) pre-embedded in the tank's outer wall, and stress data is collected through strain gauges welded to key stress areas of the tank. This real-time data is processed using a Kalman filter algorithm to eliminate noise and is used to continuously correct the state of the virtual model (such as the temperature and stress fields), ensuring a high degree of synchronization between the model's predicted values ​​and the physical entity's state.

[0021] Multi-objective optimization decision-maker: This decision-maker is based on real-time data output from the digital twin module. It abstracts the preheating process into a multi-objective optimization problem, with optimization objectives including: total preheating time. Maximum equivalent stress of storage tank shell (Calculated from a virtual model), and comprehensive energy consumption cost. (Calculated based on real-time electricity prices and fuel costs). It employs a non-dominated sorting genetic algorithm (NSGA-II) for solving the problem. The specific workflow is as follows: the algorithm continuously generates different heat source allocation and actuator control schemes, performs rapid simulations in a digital twin model, and finally outputs a set of Pareto optimal solutions. Each solution in this set represents the optimal solution in... , and An optimal trade-off among the three objectives.

[0022] Dynamic weighting factors (corresponding to claim 7): To enable the system to adapt to changes in external conditions, the optimizer introduces dynamic weights. For example, during off-peak hours at night (when electricity prices are low), the algorithm automatically increases the weight of the "lowest operating cost" objective, tending to use more electric heating; when the weather forecast predicts a cloudy day for the next day (low photovoltaic power generation, expected high electricity prices), the strategy will be adjusted in advance to increase the priority of waste heat utilization; when the ambient temperature drops sharply, the weight of "preheating time" will be appropriately increased due to increased heat loss.

[0023] Real-time controller: It takes the Pareto optimal solution set output by the multi-objective optimization decision-maker and, according to preset preferences (e.g., in any case, ...), Select a current optimal solution (not exceeding 80% of the material's allowable stress). Then, translate this solution into specific control commands, such as: "Instruct the output power of the transient heat source to increase to 70%, instruct the preheating subunit in area A to switch to radiation mode, and instruct the power in area B to decrease by 20%."

[0024] 4. Detailed description of the distributed preheating actuator array.

[0025] This array is responsible for precisely applying heat to the storage tank.

[0026] As described in claim 5, the array comprises multiple independently addressable preheating sub-units. Each sub-unit is essentially a device integrating multiple heating elements. For example, a sub-unit may include: 1) components in a hot air mode: a centrifugal fan and a set of heating wires for convective heating of the tank wall; and 2) components in a directional infrared mode: several quartz infrared heating tubes for radiative heating of a specific area.

[0027] The dynamic thermal management unit instructs each subunit to act based on real-time thermal stress risk calculated by the digital twin model. For example, when the model predicts that the stress in the lower end cap area of ​​the tank is approaching the safety threshold due to a structural abrupt change, it will instruct the subunit in that area to switch from a high-energy-density infrared mode to a milder hot air mode to reduce thermal shock. At the same time, to compensate for the overall temperature rise delay caused by the power reduction in that area, the system may instruct the subunits in the lower stress areas of the cylinder to appropriately increase their power.

[0028] 5. Example of running the adaptive preheating strategy.

[0029] The core of the strategy is the "thermal stress safety window" defined in claim 6. The upper limit of this window is not a fixed value, but a dynamically calculated stress limit. It is determined by three variables: the current material temperature (the allowable stress may change as the temperature increases), the maximum rate of temperature rise over a past period (reflecting historical thermal shocks), and the predicted future rate of temperature rise acceleration (predicting potential risks).

[0030] Example: During initial system preheating, the safety window is relatively wide, allowing for rapid temperature rise. When the temperature rise rate in a certain area is too fast, and its acceleration is predicted to be positive, the dynamic thermal management unit will narrow the "safety window" for that area. Once the predicted stress will exceed this window, the power reduction, mode switching, and neighborhood compensation operations described above will be immediately executed to proactively suppress stress risks.

[0031] 6. Detailed description of the preheating process traceability and self-learning module.

[0032] This module is key to achieving continuous system optimization.

[0033] It records all data for each preheating cycle in detail, including: the sequence of control commands for each time step, system status (temperature, stress), external conditions (electricity price, ambient temperature), and the final performance evaluation indicators.

[0034] As described in claim 9, it constructs a preheating effect evaluation function, for example:

[0035] in, , , The weighting coefficient is a dimensionless adjustment parameter. Its specific value is preset by the system developers or users based on priority preferences in actual production or initially calibrated through system self-learning.

[0036] Total heating capacity refers to the total effective heat actually absorbed by the molten salt tank and its internal medium (such as molten salt) throughout the entire preheating process. This parameter can be estimated by calculating the heat required for the overall temperature rise of the tank, or by cumulative measurement using a heat flow meter integrated into the output of the preheating actuator array. Its unit is typically megajoules or kilowatt-hours.

[0037] The term "preheating" refers to the total cost of all energy consumed to complete the entire preheating process.

[0038] The actual preheating time refers to the total time elapsed from the start of the preheating system until the molten salt storage tank reaches the preset target temperature (e.g., 400°C) at the latest measurement point. This parameter is automatically timed and recorded during each preheating process via an internal clock in the system control center.

[0039] The target preheating time refers to the total preheating time that is expected to be completed in advance, based on production plans, equipment scheduling requirements, or optimal economic analysis. This parameter is input into the system control center by the operator or the upper-level energy management system before preheating begins.

[0040] The maximum equivalent stress, as measured in practice, refers to the maximum value in the stress data sequence collected by stress sensors (such as resistance strain gauges or fiber optic strain sensors) deployed at key locations on the outer wall of the molten salt tank (such as the connection between the cylinder and the head, and the support attachment area) throughout the entire preheating process. This data is processed by a signal conditioner and then recorded by the system.

[0041] The maximum equivalent stress, as measured in practice, refers to the maximum value in the stress data sequence collected by stress sensors (such as resistance strain gauges or fiber optic strain sensors) deployed at key locations on the outer wall of the molten salt tank (such as the connection between the cylinder and the head, and the support attachment area) throughout the entire preheating process. This data is processed by a signal conditioner and then recorded by the system.

[0042] The module employs reinforcement learning algorithms (such as DeepQ-Network, DQN). The system treats each warm-up process as a "round," with the final evaluation function value F serving as the reward signal. Through iterative learning across numerous rounds, the algorithm gradually explores which control policies (i.e., state-action pairs) yield higher cumulative rewards (i.e., better overall performance) under different initial states and external conditions, thereby continuously updating and enriching its policy library. Upon the next warm-up initiation, the dynamic thermal management unit prioritizes matching and initializing from the learned best policies, resulting in progressively higher efficiency and safety with each generation.

[0043] This invention constructs an intelligent preheating system integrating perception, decision-making, execution, and learning through the precise collaborative work of the aforementioned units. Those skilled in the art can adjust and implement the specific methods based on the above detailed description and common knowledge in the field. Therefore, this invention is not limited to the specific embodiments disclosed, but covers all modifications and equivalents falling within the scope of protection defined in the claims.

[0044] The steps of the methods or algorithms described in this application can be directly embedded in hardware, software units executed by a processor, or a combination of both. Exemplarily, a storage medium can be connected to a processor so that the processor can read information from and write information to the storage medium. Optionally, the storage medium can also be integrated into the processor. Optionally, the processor and the storage medium can also be located in different components within a terminal. These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0045] Although this application has been described in conjunction with specific features and embodiments, it is obvious that various modifications and combinations can be made thereto without departing from the spirit and scope of this application. Accordingly, this specification and drawings are merely illustrative examples of this application and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Thus, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A composite preheating system for a molten salt storage tank, characterized in that, include: At least two heterogeneous heat source modules with different thermodynamic properties and response rates; A dynamic thermal management unit is configured to receive the output of the heterogeneous heat source module and execute an adaptive preheating strategy based on online inversion of the stress field of the molten salt tank structure and prediction of thermal shock risk. And a distributed preheating actuator array, which is controlled by the dynamic thermal management unit and performs multivariable coupled control based on the real-time temperature field and stress field distribution of the molten salt tank; The optimization objectives of the adaptive preheating strategy include at least the shortest preheating time, the minimum thermal stress, and the lowest operating cost. The dynamic thermal management unit solves this multi-objective optimization problem to dynamically allocate the contribution of each heterogeneous heat source module and the operating mode of the distributed preheating actuator array in real time.

2. The composite preheating system for a molten salt storage tank according to claim 1, characterized in that: The heterogeneous heat source module includes a high-power-density transient electric heat source and a high-thermal-inertia industrial process waste heat source; the power density of the transient electric heat source is not less than 50 kW / m³. 3 The response time is less than 10 seconds; the temperature fluctuation of the industrial process waste heat source is limited to within ±5% of its rated output temperature.

3. The composite preheating system for a molten salt storage tank according to claim 2, characterized in that: The dynamic thermal management unit includes: a digital twin module for constructing and updating a virtual tank model that is synchronized with the physical molten salt tank status in real time; A multi-objective optimization decision-maker, based on the virtual tank model, with preheating time... Maximum equivalent stress of storage tank shell and overall energy consumption costs As input, the Pareto optimal solution set is obtained through a non-dominated sorting genetic algorithm; A real-time controller generates control commands for the heterogeneous heat source module and the distributed preheating actuator array based on the current optimal solution output by the multi-objective optimization decision-maker.

4. The composite preheating system for a molten salt storage tank according to claim 3, characterized in that: The digital twin module acquires temperature and stress field data in real time through a distributed fiber optic sensor network and stress sensors embedded in the outer wall of the molten salt tank, and uses this data to correct the state variables of the virtual tank model using a Kalman filter algorithm.

5. The composite preheating system for a molten salt storage tank according to claim 1, characterized in that: The distributed preheating actuator array includes multiple independently addressable preheating sub-units, each of which has at least two different heat transfer modes. The heat transfer modes include a convection-based accompanying hot air mode and a radiation-based directional infrared mode. The dynamic thermal management unit dynamically switches or combines the heat transfer modes of the preheating sub-units based on the real-time thermal stress risk assessment results of different areas of the molten salt storage tank.

6. The composite preheating system for a molten salt storage tank according to claim 5, characterized in that: The adaptive preheating strategy defines a dynamically changing "thermal stress safety window". The boundary value of this window is jointly determined by the current temperature of the tank material, the historical maximum temperature rise rate, and the predicted future temperature rise acceleration. When the stress in a certain area is predicted to exceed the "thermal stress safety window", the dynamic thermal management unit instructs the preheating subunit at the corresponding location to reduce power or switch to a heat transfer mode with less thermal shock, and instructs the preheating subunits in other areas to perform power compensation.

7. The composite preheating system for a molten salt storage tank according to claim 3, characterized in that: The multi-objective optimization decision-maker introduces a dynamic weighting factor based on external conditions. This weighting factor is adaptively adjusted according to real-time grid electricity price signals, industrial process waste heat availability forecasts, and ambient temperature to change the priority of the objective of minimizing operating costs among the optimization objectives.

8. The composite preheating system for a molten salt storage tank according to claim 1, characterized in that: The system also includes a preheating process traceability and self-learning module, which records the operation sequence, system status and final effect evaluation index of the entire preheating process; through reinforcement learning algorithm, the control parameters in the adaptive preheating strategy are iteratively optimized, so that the preheating efficiency and stress control effect of the system are continuously improved after multiple preheating cycles.

9. The composite preheating system for a molten salt storage tank according to claim 8, characterized in that: The preheating process traceability and self-learning module constructs a preheating effect evaluation function, which comprehensively considers the deviation between the actual preheating time and the target time, the ratio of the measured maximum stress to the allowable stress, and the comprehensive cost per unit preheating temperature rise. The system updates its strategy library by comparing the evaluation function values ​​under different control strategies.

10. The composite preheating system for a molten salt storage tank according to claim 1, characterized in that: The heterogeneous heat source module integrates a stepped heating subsystem for deep recovery of low-temperature waste heat. This subsystem uses a non-azeotropic mixed working fluid as an intermediate heat transfer medium. By adjusting the concentration of the working fluid components, the evaporation process is nonlinearly matched with the cooling curve of the industrial process waste heat source, and the condensation process is nonlinearly matched with the preheating requirements of the molten salt storage tank.

Images