Entropy production constrained digital twinning optimization system, method and device, storage medium and program method

Abstract

The invention provides an entropy production constrained digital twinning optimization system, method and device, a storage medium and a program method. The entropy yield is directly used as a control target, and errors caused by intermediate parameter conversion are eliminated. Different regulation and control strategy effects are predicted through digital twinborn simulation, and dynamic optimization is achieved. And the response time is shortened to a second level through edge calculation and real-time feedback. And real-time sensing and accurate suppression of the loss of the thermodynamic system are realized.

Description

Technical Field

[0001] This application relates to the field of energy system control technology, and in particular to a digital twin with entropy production constraints. Optimize systems, methods, devices, storage media, and program approaches. Background Technology

[0002] In the field of intelligent control of energy systems, existing technologies mainly improve system efficiency by monitoring system parameters and adjusting them based on preset rules. Typical solutions include monitoring and adjusting the parameters of new energy power generation equipment, as well as evaluating and optimizing the heat exchange efficiency of air-cooled systems.

[0003] However, these traditional methods have three significant drawbacks: First, system monitoring and dynamic control are isolated, making closed-loop control impossible; second, control strategies rely on empirical rules, making them ill-suited to complex operating conditions; and finally, the system response exhibits a significant delay, failing to promptly capture energy losses caused by transient changes in operating conditions. Particularly in handling irreversible thermodynamic losses, existing technologies often focus only on indirect parameters such as temperature and pressure, failing to directly use entropy production rate as the control variable, leading to system... Damage positioning is lagging and control precision is insufficient. Summary of the Invention

[0004] To address the aforementioned technical problems, this application provides a digital twin with entropy production constraints. Optimize systems, methods, devices, storage media, and program approaches.

[0005] Firstly, embodiments of this application provide a digital twin constrained by entropy production. Optimize the system, including:

[0006] The entropy production sensing terminal is used to calculate the real-time entropy production rate of each subsystem based on the multi-dimensional physical field data of each subsystem in the target system collected in real time, and to generate an entropy production rate cloud map of each subsystem.

[0007] The entropy production feedback module is used to determine the entropy production change rate and cumulative entropy production change rate of each subsystem based on the real-time entropy production rate of each subsystem, and to obtain the real-time control quantity of each subsystem based on the entropy production change rate and cumulative entropy production change rate of each subsystem.

[0008] The flow topology optimization module is used to determine the high entropy production area in each subsystem based on the numerical distribution of the entropy production cloud map of each subsystem and the preset threshold rules, and to match the execution device responsible for regulating each high entropy production area according to the preset energy flow path and device topology relationship.

[0009] The digital twin construction module is used to simulate and output the entropy production rate change trend of each subsystem under the action of the real-time control quantity, as well as the key performance parameters of the execution device responsible for controlling each high entropy production region in each subsystem, taking the real-time control quantity of each subsystem as input.

[0010] The decision module is used to determine the actual control quantity of the execution device in each subsystem that matches the high-entropy-producing region, based on the real-time control quantity of each subsystem, the entropy production rate change trend of each subsystem, and the key performance parameters of the execution device responsible for controlling each high-entropy-producing region in each subsystem, on the premise of meeting the equipment safety threshold and / or system stability boundary of each subsystem, and to send the actual control quantity of the execution device to the corresponding execution mechanism, so as to minimize the real-time entropy production rate of each subsystem.

[0011] In some embodiments, the decision-making module includes at least an actual control quantity determination unit, the actual control quantity determination unit being used for:

[0012] Based on the entropy production rate change trend of each subsystem, determine the control coefficient of each subsystem;

[0013] Based on the key performance parameters of the execution equipment responsible for regulating each high-entropy production region in each subsystem, determine the adjustable margin of the execution equipment in each high-entropy production region.

[0014] Based on the premise of satisfying the equipment safety threshold and / or system stability boundary of each subsystem, the actual control amount of the execution equipment in each subsystem that matches the high-entropy production region is determined according to the real-time control amount of each subsystem, the control coefficient of each subsystem, and the adjustable margin of the execution equipment in each high-entropy production region.

[0015] In some embodiments, the step of determining the adjustable margin of the execution device in each high-entropy production region includes:

[0016] Based on the entropy production rate cloud map of each subsystem, determine the entropy production level corresponding to the high entropy production region in each subsystem;

[0017] Based on the entropy production level corresponding to the high entropy production region in each subsystem, determine the maximum adjustable margin of the execution equipment in each high entropy production region;

[0018] Based on the operating parameters of the execution devices in each high-entropy production region and the maximum adjustable margin of the execution devices in each high-entropy production region, and on the premise of satisfying the safety threshold of each execution device, the adjustable margin of the execution devices in each high-entropy production region is determined.

[0019] In some embodiments, the entropy generation feedback module includes at least an adaptive gain unit, the adaptive gain unit being used for:

[0020] The entropy production rate and entropy production rate gain of each subsystem are determined by the derivative element in the adaptive gain PID control method.

[0021] The cumulative entropy productivity and cumulative entropy productivity gain of each subsystem are determined by the integral element in the adaptive gain PID control method.

[0022] The real-time control quantity of each subsystem is obtained based on the entropy production change rate, entropy production change rate gain, cumulative entropy production rate, and cumulative entropy production rate gain of each subsystem.

[0023] In some embodiments, the The flow topology optimization module includes at least an execution device determination unit, which is used for:

[0024] When the entropy production change rate of a certain subsystem is greater than 0, the acceleration fan responsible for regulating each high entropy production region in that subsystem will be used as the execution device for regulating each high entropy production region in that subsystem.

[0025] If the cumulative entropy production rate of a certain subsystem exceeds a preset cumulative threshold, the cleaning equipment responsible for regulating each high-entropy production area in that subsystem will be used as the execution equipment for regulating each high-entropy production area in that subsystem.

[0026] In some embodiments, the adaptive gain unit is further configured to:

[0027] The real-time load rate of each subsystem is determined based on the maximum rated power and real-time output power of each subsystem.

[0028] The real-time load change rate of each subsystem is determined based on the real-time load rate of each subsystem.

[0029] If the real-time load change rate of the first subsystem exceeds the preset threshold load change rate, the reward value of the first subsystem is generated based on the real-time entropy production rate, temperature gradient, vorticity and pressure difference of the first subsystem through the preset reward function.

[0030] Based on the range of the reward value of the first subsystem, determine the correction direction and adjustment range of the entropy production change rate gain and the cumulative entropy production change rate gain of the first subsystem.

[0031] Based on the correction direction and the adjustment magnitude, adjust the entropy production rate change gain and the cumulative entropy production rate change gain of the first subsystem to maximize the reward value of the first subsystem.

[0032] The entropy change rate gain and cumulative entropy deviation gain corresponding to the maximum reward value of the first subsystem are determined as the entropy change rate gain and cumulative entropy deviation gain of the first subsystem.

[0033] Secondly, embodiments of this application provide a digital twin with entropy production constraints. Optimization methods include:

[0034] Based on the multi-dimensional physical field data of each subsystem in the target system collected in real time, the real-time entropy production rate of each subsystem is calculated, and an entropy production rate cloud map of each subsystem is generated.

[0035] Based on the real-time entropy production rate of each subsystem, determine the entropy production change rate and cumulative entropy production change rate of each subsystem, and obtain the real-time control quantity of each subsystem based on the entropy production change rate and cumulative entropy production change rate of each subsystem.

[0036] Based on the numerical distribution of the entropy productivity cloud map of each subsystem and the preset threshold rules, the high entropy production area of ​​each subsystem is determined, and the execution device responsible for regulating each high entropy production area is matched according to the preset energy flow path and device topology relationship.

[0037] Using the real-time control quantity of each subsystem as input, the simulation outputs the entropy production rate change trend of each subsystem under the action of the real-time control quantity, as well as the key performance parameters of the execution equipment responsible for controlling each high entropy production region in each subsystem.

[0038] Based on the real-time control amount of each subsystem, and through the entropy production rate change trend of each subsystem and the key performance parameters of the execution equipment responsible for controlling each high-entropy production region in each subsystem, and on the premise of meeting the equipment safety threshold and / or system stability boundary of each subsystem, the actual control amount of the execution equipment matching the high-entropy production region in each subsystem is determined, and the actual control amount of the execution equipment is sent to the corresponding execution mechanism to minimize the real-time entropy production rate of each subsystem.

[0039] Thirdly, embodiments of this application provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the digital twin described above regarding entropy production constraints. Optimization methods.

[0040] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the digital twin described above regarding entropy production constraints. Optimization methods.

[0041] Fifthly, embodiments of this application provide a computer program product, which includes a computer program that, when executed by a processor, implements the digital twin described above regarding entropy production constraints. Optimization methods.

[0042] Compared with the prior art, this application has the following advantages:

[0043] This application's embodiments directly use entropy production rate as the control objective, eliminating errors caused by intermediate parameter transformations. Dynamic optimization is achieved by predicting the effects of different control strategies through digital twin simulation. Response time is reduced to the second level through edge computing and real-time feedback, realizing the thermal system... Real-time perception and precise suppression of entropy buildup. Under high-temperature conditions, the system can quickly identify localized eddy current regions and adjust the corresponding fan speed accordingly to prevent entropy accumulation. In equipment scaling scenarios, it promptly triggers cleaning commands to maintain a low-entropy state. System stability is ensured through digital twin simulation, avoiding the risk of equipment overload caused by excessive regulation. Attached Figure Description

[0044] Figure 1 This is a digital twin of entropy production constraints provided in one embodiment of this application. Optimize the system's structural diagram;

[0045] Figure 2 This is a digital twin of entropy production constraints provided in one embodiment of this application. An exemplary structural diagram of the optimization system;

[0046] Figure 3 This is a digital twin of entropy production constraints provided in one embodiment of this application. Optimize the flowchart of the method;

[0047] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0048] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0049] In existing technologies, intelligent control of energy systems has long relied on indirect parameters such as temperature and pressure for regulation, making it difficult to directly suppress irreversible thermodynamic losses. Traditional methods train models using historical data, resulting in response delays exceeding ten minutes and failing to capture transient entropy production changes. For example, when an air-cooled system employs a uniform speed regulation strategy, the entropy increase caused by local eddies cannot be eliminated in a timely manner, leading to… Accumulated losses. Under high-temperature operating conditions, a thermal power unit experienced a sharp increase in local entropy production rate due to delayed control, leading to a decrease in equipment efficiency.

[0050] To address the aforementioned issues, the applicant found that existing regulatory mechanisms do not directly control the entropy production rate, leading to… The effect of loss suppression is limited. By analyzing the energy flow path of the thermal system, a dynamic closed-loop control is constructed by combining real-time entropy production feedback with equipment topology optimization. Furthermore, considering that digital twin technology can simulate the effect of controllable quantities, an adaptive control strategy based on entropy production cloud maps is designed to achieve precise equipment-level control.

[0051] Reference Figure 1 This illustrates a digital twin with entropy production constraints proposed in an embodiment of this application. Optimize the system, including:

[0052] The entropy production sensing terminal 10 is used to calculate the real-time entropy production rate of each subsystem based on the multi-dimensional physical field data of each subsystem in the target system collected in real time, and to generate an entropy production rate cloud map of each subsystem.

[0053] The entropy production feedback module 20 is used to determine the entropy production change rate and cumulative entropy production change rate of each subsystem based on the real-time entropy production rate of each subsystem, and to obtain the real-time control quantity of each subsystem based on the entropy production change rate and cumulative entropy production change rate of each subsystem.

[0054] The flow topology optimization module 30 is used to determine the high entropy production area in each subsystem based on the numerical distribution of the entropy production rate cloud map of each subsystem and the preset threshold rules, and to match the execution device responsible for regulating each high entropy production area according to the preset energy flow path and device topology relationship.

[0055] The digital twin construction module 40 is used to simulate and output the entropy production rate change trend of each subsystem under the action of the real-time control quantity, as well as the key performance parameters of the execution equipment responsible for controlling each high-entropy production region in each subsystem, taking the real-time control quantity of each subsystem as input.

[0056] The decision module 50 is used to determine the actual control quantity of the execution equipment in each subsystem that matches the high-entropy-production region, based on the real-time control quantity of each subsystem, the entropy production rate change trend of each subsystem, and the key performance parameters of the execution equipment responsible for controlling each high-entropy-production region in each subsystem, on the premise of meeting the equipment safety threshold and / or system stability boundary of each subsystem, and to send the actual control quantity of the execution equipment to the corresponding execution mechanism, so as to minimize the real-time entropy production rate of each subsystem.

[0057] Among them, the entropy production sensing terminal 10 refers to the sensor network and edge computing unit deployed in the subsystem, which can be implemented using distributed temperature and flow rate sensors in conjunction with an embedded processor to capture local thermodynamic states. The entropy production feedback module 20 refers to the computing unit based on an adaptive control algorithm, which can be implemented using a PID controller combined with a reinforcement learning weight update mechanism to quantify control requirements. The topology optimization module 30 refers to the database and matching engine for storage device topology relationships, which can be implemented using a graph database and path planning algorithm to establish a mapping between entropy-generating hotspots and execution devices. The digital twin construction module 40 refers to a virtual image of the physical system, which can be constructed using multiphysics simulation software to predict control effects. The decision module 50 refers to an optimization solver under safety constraints, which can be implemented using a constrained linear programming model to generate executable device instructions.

[0058] Specifically, during system operation, sensors in each subsystem collect multi-dimensional data such as temperature and velocity fields in real time. The entropy production sensing terminal 10 calculates the entropy production rate of each grid cell using thermodynamic formulas and generates a visual cloud map. The entropy production feedback module 20 analyzes the cloud map data, calculates the current rate of change of entropy production and its historical cumulative value, and generates preliminary control parameters through an adaptive algorithm. The flow topology optimization module 30 identifies regions exceeding a threshold in the cloud map and determines the specific fan or valve responsible for regulation based on preset equipment topology relationships. After receiving the regulation input, the digital twin simulates the entropy production distribution changes after executing the equipment action and predicts the changing trends of key performance parameters. The decision module 50 integrates equipment safety thresholds and system stability boundaries, corrects the initial regulation input, generates the final execution command, and sends it to the field equipment. The entire process forms a closed-loop control, ensuring that the entropy production rate continuously converges to its minimum value.

[0059] For example, refer to Figure 2 The entropy production sensing terminal 10, acting as the system's "sensing antennae," collaborates with the sensor network (such as temperature and flow rate sensors) and edge computing units deployed within the subsystems. It collects multi-dimensional physical field data (such as temperature and velocity fields) from each subsystem in real time and calculates the real-time entropy production rate (i.e., real-time S_gen data) of each subsystem based on thermodynamic formulas, generating a visualized entropy production rate cloud map. Edge computing nodes input the preprocessed data into the entropy production feedback loop (i.e., the entropy production feedback module 20) to obtain the real-time control parameters for each subsystem. The virtual entropy production mapping model receives the entropy production rate cloud map from the entropy production sensing terminal 10, constructs a digital twin, and uses the output of the virtual entropy production mapping model as input to simulate the entropy production rate change trend of the subsystem under the influence of real-time control parameters. Simultaneously, it predicts the key performance parameters (such as energy consumption and wear rate) of the control-responsible execution equipment (such as fans and valves); the entropy flow topology optimizer (i.e.... The flow topology optimization module 30, combining the simulation results of the digital twin, identifies high-entropy-producing regions in the cloud map that exceed the threshold based on preset equipment topology relationships (such as the connection logic of pipes and valves) and threshold rules (the criteria for judging high and low entropy production). It then matches the execution equipment responsible for regulating the region (such as a fan corresponding to a high-entropy-producing region). The decision engine (i.e., the decision module 50) receives the "equipment-region matching results" from the entropy flow topology optimizer and the "regulation trend prediction" from the digital twin. Using equipment safety thresholds (such as the upper limit of fan speed) and system stability boundaries (such as the pressure fluctuation range) as constraints, it generates the actual regulation amount of the execution equipment (such as increasing the fan speed by 5% and adjusting the valve opening to 30%) through optimization algorithms, generates regulation commands, and sends them to each execution mechanism. After receiving the control instructions output by the decision engine, the actuator performs physical-level control actions (such as changing the speed of the variable frequency fan or adjusting the opening of the electric valve), directly intervening in the physical field state of the system (such as flow velocity and temperature distribution). After the actuator takes action, the new physical field state of the system is collected again by the entropy production sensing terminal 10, and the data flows back to the edge computing node, re-entering the "entropy production feedback loop", forming a closed-loop control of "sensing → analysis → decision → execution → re-sensing", continuously driving the real-time entropy production rate of the system to converge toward the theoretical minimum.

[0060] This application's embodiments directly use entropy production rate as the control objective, eliminating errors caused by intermediate parameter transformations. Dynamic optimization is achieved by predicting the effects of different control strategies through digital twin simulation. Response time is reduced to the second level through edge computing and real-time feedback, realizing the thermal system... Real-time perception and precise suppression of entropy buildup. Under high-temperature conditions, the system can quickly identify localized eddy current regions and adjust the corresponding fan speed accordingly to prevent entropy accumulation. In equipment scaling scenarios, it promptly triggers cleaning commands to maintain a low-entropy state. System stability is ensured through digital twin simulation, avoiding the risk of equipment overload caused by excessive regulation.

[0061] In some embodiments, the decision module 50 includes at least an actual control quantity determination unit, which is used for:

[0062] The control coefficient for each subsystem is determined based on the trend of entropy production rate change in each subsystem.

[0063] Based on the key performance parameters of the execution equipment responsible for regulating each high-entropy production region in each subsystem, determine the adjustable margin of the execution equipment in each high-entropy production region.

[0064] Based on the premise of meeting the equipment safety threshold and / or system stability boundary of each subsystem, the actual control amount of the execution equipment in each subsystem that matches the high-entropy production region is determined according to the real-time control amount of each subsystem, the control coefficient of each subsystem, and the adjustable margin of the execution equipment in each high-entropy production region.

[0065] The control coefficient is a parameter reflecting the influence of the subsystem's entropy productivity change trend on the control amount. Specifically, it can be calculated by dividing the slope of the entropy productivity change curve by a preset benchmark value, and is used to dynamically adjust the control intensity. The adjustable margin refers to the maximum allowable control range of the executing equipment within its safe operating range. Specifically, it can be calculated based on the equipment's rated power, the difference between current operating parameters and a safety threshold, and is used to constrain the actual control amount from exceeding the equipment's carrying capacity.

[0066] Specifically, when calculating the control coefficient based on the trend of entropy productivity change, a trend model can be established based on time series data. For example, a sliding window algorithm can be used to calculate the gradient of entropy productivity change within the most recent time window. Key performance parameters of the executing equipment include, but are not limited to, rated power, response time, and adjustment accuracy. These parameters, together with the current operating state of the equipment, determine the upper limit of the adjustable margin. When determining the actual control amount, the real-time control amount is multiplied by the control coefficient to obtain the basic control amount. Then, the basic control amount is limited according to the adjustable margin to ensure that the control command is within the equipment safety threshold and system stability boundary.

[0067] In this embodiment, by dynamically calculating the control coefficient and adjustable margin, real-time matching between control parameters and system state is achieved, solving the problems of control lag or equipment overload caused by empirical rules. It can dynamically adjust the control intensity according to the trend of entropy production rate changes, and precisely constrain the control range by combining the real-time operating status of the equipment. This improves entropy production suppression efficiency and avoids operational risks caused by the control amount exceeding the equipment's carrying capacity, achieving a dual optimization of safety and control accuracy.

[0068] In some embodiments, the step of determining the adjustable margin of the execution device in each high-entropy production region includes:

[0069] Based on the entropy production rate cloud map of each subsystem, determine the entropy production level corresponding to the high entropy production region in each subsystem.

[0070] Based on the entropy production level corresponding to the high-entropy production region in each subsystem, determine the maximum adjustable margin of the execution equipment in each high-entropy production region.

[0071] Based on the operating parameters of the execution equipment in each high-entropy production region, and the maximum adjustable margin of the execution equipment in each high-entropy production region, the adjustable margin of the execution equipment in each high-entropy production region is determined on the premise of meeting the safety threshold of each execution equipment.

[0072] Among them, entropy production level refers to the quantitative index of thermodynamic irreversibility divided by the numerical distribution range in the entropy production rate cloud map. Specifically, it can be achieved by dividing the cloud map into multiple numerical intervals using a gradient threshold segmentation algorithm, with each interval corresponding to a different control priority. Maximum adjustable margin refers to the maximum allowable adjustment range of the executing equipment under safe operation conditions. This can be achieved by matching and querying the mapping relationship between the equipment performance curve and the entropy production level; a higher entropy production level corresponds to a larger adjustable range. Operating parameters include, but are not limited to, real-time monitoring data such as temperature, pressure, and vibration amplitude. These can be obtained through a combination of sensor acquisition and edge computing node preprocessing, used to dynamically correct the upper limit of the adjustable margin. Safety threshold refers to the limitations on the executing equipment in terms of mechanical strength and thermal stress resistance. Specifically, it can be obtained by multiplying the rated parameters provided by the equipment manufacturer by the correction coefficient under real-time operating conditions to obtain the dynamic threshold.

[0073] Specifically, the entropy productivity cloud map is generated by fusing multi-dimensional physical field data. A region segmentation algorithm is then used to identify regions with entropy productivity exceeding a benchmark value as high-entropy productivity regions. Each high-entropy productivity region is assigned a corresponding entropy productivity level based on the range in which its entropy productivity value falls; for example, 500-600 W / (m³)

[0074] ·K) is classified as Level 1, and 600-700W / (m3·K) is classified as Level 2. The maximum adjustable margin of the actuator is determined by querying a pre-established entropy production level-control capability mapping table, which is generated based on the performance test data of the equipment under laboratory conditions. When determining the actual adjustable margin, the current operating parameters of the actuator are input into the dynamic constraint model. For example, when the fan bearing temperature exceeds the set threshold, its adjustable margin is reduced proportionally. The final output adjustable margin is the smaller value between the maximum adjustable margin and the value calculated by the dynamic constraint model, ensuring that the control operation is always within the equipment's safety boundary.

[0075] This application's embodiments establish a quantitative correlation between entropy production levels and equipment performance, and dynamically correct the control boundary using real-time operating parameters, achieving optimal matching between equipment control capabilities and thermodynamic states. This effectively solves the problem of the disconnect between equipment control margin settings and real-time entropy production states. Through the synergistic effect of entropy production level classification and dynamic constraint models, control efficiency is maximized while ensuring equipment safety, avoiding the risks of insufficient control or equipment overload caused by improper margin settings in traditional methods.

[0076] In some embodiments, the entropy generation feedback module 20 includes at least an adaptive gain unit, which is used for:

[0077] The entropy production rate and entropy production rate gain of each subsystem are determined by the derivative element in the adaptive gain PID control method.

[0078] The cumulative entropy production rate and cumulative entropy production rate gain of each subsystem are determined by the integral element in the adaptive gain PID control method.

[0079] The real-time control parameters for each subsystem are obtained based on the rate of change of entropy production, the gain of the rate of change of entropy production, the cumulative entropy production rate, and the gain of the cumulative entropy production rate.

[0080] In the adaptive gain PID control method, the derivative element refers to calculating the entropy production change rate using the first derivative of the real-time entropy production rate, and then amplifying or suppressing the change trend by combining it with a dynamically adjusted gain coefficient. Specifically, the real-time load rate can be used as the input parameter for gain adjustment. The integral element refers to integrating the deviation of the entropy production rate accumulated over time. The cumulative entropy production rate gain is dynamically corrected based on the subsystem's operating state, which can be achieved by optimizing the integral coefficients online using a reinforcement learning model. The adjustment direction of the entropy production change rate gain and the cumulative entropy production rate gain is determined by comparing the real-time load change rate with a preset threshold. Specifically, a gain correction mechanism is triggered when the load fluctuation exceeds the threshold.

[0081] Specifically, during system operation, the derivative term continuously monitors the instantaneous rate of change of entropy production rate. When a rapid increase in entropy production rate in a certain area is detected, the signal strength is amplified by dynamically adjusting the gain coefficient. The integral term simultaneously calculates the cumulative amount of historical entropy production deviation in that area and optimizes the integral gain online based on the load rate change trend. For example, when a sudden increase in the subsystem load rate leads to a surge in entropy production, the integral gain is automatically increased to accelerate the elimination of the accumulated deviation. Through the synergistic effect of the derivative and integral terms, the real-time control quantity can simultaneously consider transient response speed and steady-state adjustment accuracy, avoiding the over-adjustment or under-adjustment phenomena caused by traditional fixed-parameter PID control.

[0082] This application's embodiments introduce load rate as the basis for gain adjustment, enabling control parameters to automatically optimize according to the system's operating state. For example, during peak shaving of thermal power units, when the load change rate reaches a critical value, the gain coefficient will be dynamically corrected according to a preset reinforcement learning strategy, shortening the adjustment convergence time by 40% compared to conventional PID control. This solves the response delay problem caused by fixed gain in traditional entropy production control, achieving dynamic parameter matching under complex operating conditions. In the air-cooled condenser operation scenario, when a sudden change in ambient temperature causes a surge in local entropy production, the adaptive gain mechanism can shorten the generation time of control commands to the second level, while controlling the fluctuation range of entropy production rate within a safe threshold range, effectively avoiding mechanical stress damage to equipment caused by over-adjustment.

[0083] In some embodiments, The flow topology optimization module 30 includes at least an execution device determination unit, which is used for:

[0084] When the entropy production rate of a certain subsystem is greater than 0, the acceleration fan responsible for regulating each high-entropy production region in that subsystem will be used as the execution device for regulating each high-entropy production region in that subsystem.

[0085] If the cumulative entropy production rate of a certain subsystem exceeds the preset cumulative threshold, the cleaning equipment responsible for regulating each high-entropy production area in that subsystem will be used as the execution equipment for regulating each high-entropy production area in that subsystem.

[0086] Among these, an entropy production change rate greater than 0 indicates that the entropy production rate of the subsystem is increasing. This can be achieved by collecting physical field data in real time and calculating the difference in entropy production rates between adjacent time points. This indicator reflects the current trend of the system's thermodynamic irreversibility. A cumulative entropy production rate exceeding a preset cumulative threshold means that the total entropy production of the subsystem over a period of time reaches a critical value. This can be achieved by integrating and summing the entropy production rates at each time point. This indicator is used to assess the accumulation of energy losses during long-term system operation. An accelerating fan refers to a device that changes the airflow distribution by adjusting its rotational speed. A variable frequency speed-regulating fan can be used to suppress the increase in entropy production caused by local eddies. A cleaning device refers to a device that removes deposits from heat exchange surfaces. A high-pressure water jet device can be used to eliminate heat transfer degradation caused by scaling.

[0087] Specifically, when a positive entropy production rate is detected in the subsystem, it indicates that the thermodynamically irreversible process in that region is intensifying. At this point, adjusting the airflow speed by accelerating the fan can effectively disrupt the vortex structure and reduce flow entropy production. When the cumulative entropy production rate exceeds a preset threshold, it indicates that long-term energy loss has accumulated in that region. Activating the cleaning equipment to remove surface deposits at this time can restore heat transfer efficiency and reduce heat transfer entropy production. The switching logic between the two actuators is based on a dynamic comparison of real-time monitoring data and preset thresholds, ensuring that the control measures match the entropy production mechanism.

[0088] This application establishes selection criteria for execution equipment by quantifying two dimensions: the rate of change of entropy production and the cumulative entropy production rate. This achieves a precise correspondence between control measures and the root causes of entropy production. It rapidly matches the optimal execution equipment for entropy production problems of different causes, promptly suppressing flow losses in the initial stage of eddy currents and effectively removing deposits when scale accumulation reaches a critical value, thereby maintaining a low-entropy production state of the system under dynamic operating conditions.

[0089] In some embodiments, the adaptive gain unit is further configured to:

[0090] The real-time load rate of each subsystem is determined based on its maximum rated power and real-time output power.

[0091] The real-time load change rate of each subsystem is determined based on the real-time load rate of each subsystem.

[0092] If the real-time load change rate of the first subsystem exceeds the preset threshold load change rate, the reward value of the first subsystem is generated based on the real-time entropy production rate, temperature gradient, vorticity, and pressure difference of the first subsystem through a preset reward function.

[0093] Based on the range of the reward value of the first subsystem, determine the correction direction and adjustment range of the entropy production change rate gain and the cumulative entropy production change rate gain of the first subsystem.

[0094] Based on the correction direction and adjustment magnitude, adjust the entropy production change rate gain and cumulative entropy production change rate gain of the first subsystem to maximize the reward value of the first subsystem.

[0095] The entropy production change rate gain and cumulative entropy production deviation gain corresponding to the maximum reward value of the first subsystem are determined as the entropy production change rate gain and cumulative entropy production deviation gain of the first subsystem.

[0096] The real-time load factor refers to the ratio of current output power to maximum rated power. This can be achieved by using power sensors to collect real-time power data and calculate the ratio, used to assess the current load status of the subsystem. The real-time load change rate refers to the change in load factor per unit time, calculated using a time-series differencing algorithm, used to identify the intensity of load fluctuations. The reward function is a mathematical model that uses entropy production rate, temperature gradient, vorticity, and pressure difference as input variables to generate a quantitative evaluation value. It can be constructed using linear weighting or neural network models, used to comprehensively evaluate the impact of control strategies on system stability.

[0097] Specifically, when the load change rate of a subsystem exceeds a preset threshold, the system automatically triggers a gain parameter adjustment mechanism. Real-time entropy production rate, temperature gradient, vorticity distribution, and pressure difference data of the subsystem are collected and input into a pre-trained reward function model for multi-dimensional evaluation. Based on the numerical range of the output reward value, the adjustment direction and magnitude of the gain parameters are determined. For example, when the reward value is in a low range, the correction magnitude of the entropy production change rate gain is automatically increased, while the adjustment step size of the cumulative entropy production deviation gain is decreased. Through an iterative optimization process, the combination of gain parameters that maximizes the reward function is ultimately selected, ensuring that the control strategy reduces entropy production while maintaining stable system operation.

[0098] This application's embodiments utilize a dynamic gain adjustment mechanism triggered by load changes, combined with a reward function constructed from multi-physics parameters, to achieve precise matching between control parameters and real-time operating conditions. An online optimization mechanism ensures the gain parameters are always in an optimal state. This effectively solves the control lag problem caused by fixed gain parameters in traditional PID control, maintaining stable entropy production suppression even under rapidly fluctuating load conditions. By introducing a reward function constructed from multi-physics parameters, the gain adjustment process ensures that thermodynamic performance and equipment safety are considered, avoiding system instability risks caused by optimizing a single indicator. The dynamic gain adjustment mechanism significantly improves the system's adaptability to changing operating conditions, achieving synergistic optimization of entropy production rate control accuracy and system stability.

[0099] Secondly, referring to Figure 3 This application provides a digital twin with entropy production constraints. Optimization methods include:

[0100] S10. Based on the multi-dimensional physical field data of each subsystem in the target system collected in real time, calculate the real-time entropy production rate of each subsystem and generate an entropy production rate cloud map of each subsystem.

[0101] S20. Based on the real-time entropy production rate of each subsystem, determine the entropy production change rate and cumulative entropy production change rate of each subsystem, and obtain the real-time control quantity of each subsystem based on the entropy production change rate and cumulative entropy production change rate of each subsystem.

[0102] S30. Based on the numerical distribution of the entropy production rate cloud map of each subsystem and the preset threshold rules, determine the high entropy production area in each subsystem, and match the execution equipment responsible for regulating each high entropy production area according to the preset energy flow path and equipment topology relationship.

[0103] S40. Using the real-time control quantity of each subsystem as input, simulate and output the entropy production rate change trend of each subsystem under the action of the real-time control quantity, as well as the key performance parameters of the execution equipment responsible for controlling each high-entropy production region in each subsystem.

[0104] S50. Based on the real-time control quantity of each subsystem, and through the entropy production rate change trend of each subsystem and the key performance parameters of the execution equipment responsible for controlling each high-entropy production region in each subsystem, and on the premise of meeting the equipment safety threshold and / or system stability boundary of each subsystem, determine the actual control quantity of the execution equipment matching the high-entropy production region in each subsystem, and send the actual control quantity of the execution equipment to the corresponding execution mechanism so as to minimize the real-time entropy production rate of each subsystem.

[0105] In some embodiments, it also includes:

[0106] The control coefficient for each subsystem is determined based on the trend of entropy production rate change in each subsystem.

[0107] Based on the key performance parameters of the execution equipment responsible for regulating each high-entropy production region in each subsystem, determine the adjustable margin of the execution equipment in each high-entropy production region.

[0108] Based on the premise of meeting the equipment safety threshold and / or system stability boundary of each subsystem, the actual control amount of the execution equipment in each subsystem that matches the high-entropy production region is determined according to the real-time control amount of each subsystem, the control coefficient of each subsystem, and the adjustable margin of the execution equipment in each high-entropy production region.

[0109] In some embodiments, the step of determining the adjustable margin of the execution device in each high-entropy production region includes:

[0110] Based on the entropy production rate cloud map of each subsystem, determine the entropy production level corresponding to the high entropy production region in each subsystem.

[0111] Based on the entropy production level corresponding to the high-entropy production region in each subsystem, determine the maximum adjustable margin of the execution equipment in each high-entropy production region.

[0112] Based on the operating parameters of the execution equipment in each high-entropy production region, and the maximum adjustable margin of the execution equipment in each high-entropy production region, the adjustable margin of the execution equipment in each high-entropy production region is determined on the premise of meeting the safety threshold of each execution equipment.

[0113] In some embodiments, the process of obtaining the real-time control parameters for each subsystem includes:

[0114] The entropy production rate and entropy production rate gain of each subsystem are determined by the derivative element in the adaptive gain PID control method.

[0115] The cumulative entropy production rate and cumulative entropy production rate gain of each subsystem are determined by the integral element in the adaptive gain PID control method.

[0116] The real-time control parameters for each subsystem are obtained based on the rate of change of entropy production, the gain of the rate of change of entropy production, the cumulative entropy production rate, and the gain of the cumulative entropy production rate.

[0117] In some embodiments, when the entropy production change rate of a certain subsystem is greater than 0, the acceleration fan responsible for regulating each high-entropy production region in the subsystem is used as the execution device for regulating each high-entropy production region in the subsystem.

[0118] If the cumulative entropy production rate of a certain subsystem exceeds the preset cumulative threshold, the cleaning equipment responsible for regulating each high-entropy production area in that subsystem will be used as the execution equipment for regulating each high-entropy production area in that subsystem.

[0119] In some embodiments, the process of determining the entropy production rate change gain and the cumulative entropy production deviation gain includes:

[0120] The real-time load rate of each subsystem is determined based on its maximum rated power and real-time output power.

[0121] The real-time load change rate of each subsystem is determined based on the real-time load rate of each subsystem.

[0122] If the real-time load change rate of the first subsystem exceeds the preset threshold load change rate, the reward value of the first subsystem is generated based on the real-time entropy production rate, temperature gradient, vorticity, and pressure difference of the first subsystem through a preset reward function.

[0123] Based on the range of the reward value of the first subsystem, determine the correction direction and adjustment range of the entropy production change rate gain and the cumulative entropy production change rate gain of the first subsystem.

[0124] Based on the correction direction and adjustment magnitude, adjust the entropy production change rate gain and cumulative entropy production change rate gain of the first subsystem to maximize the reward value of the first subsystem.

[0125] The entropy production change rate gain and cumulative entropy production deviation gain corresponding to the maximum reward value of the first subsystem are determined as the entropy production change rate gain and cumulative entropy production deviation gain of the first subsystem.

[0126] Thirdly, referring to Figure 4 This application also provides an electronic device, including:

[0127] processor.

[0128] Memory is used to store processor-executable instructions.

[0129] The processor is configured to execute instructions to implement a digital twin of any entropy production constraint. Optimization methods.

[0130] In this embodiment, the computer device includes a processor, memory, and network interface connected via a system bus.

[0131] The computer device's processor provides computational and control capabilities. Its memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs in the non-volatile storage media. The computer device's database stores data samples. Its network interface is used to communicate with external terminals via a network. When the computer program is executed by the processor, it implements a digital twin to achieve any entropy production constraint. Optimization methods.

[0132] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0133] Fourthly, embodiments of this application also provide a computer-readable storage medium, which, when the instructions in the computer-readable storage medium are executed by the processor of a terminal, enables the terminal to execute a digital twin subject to any entropy production constraint. Optimization methods.

[0134] The aforementioned computer-readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory, electrically erasable programmable read-only memory, erasable programmable read-only memory, programmable read-only memory, read-only memory, magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0135] Optionally, a readable storage medium can be coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Alternatively, the readable storage medium can be an integral part of the processor. Both the processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components within the device.

[0136] Fifthly, embodiments of this application also provide a computer program product, which includes a computer program that, when executed by a processor, implements a digital twin of any entropy production constraint. Optimization methods.

[0137] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0138] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0139] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0140] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment 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.

[0141] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0142] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

[0143] As the method embodiments are basically similar to the system embodiments, the description is relatively simple, and relevant parts can be found in the description of the system embodiments.

[0144] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0145] The above describes a digital twin with entropy production constraints provided in this application. The optimization system, method, device, storage medium, and program method are described in detail. Specific examples are used to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A digital twin with entropy production constraints The optimized system is characterized by, include: The entropy production sensing terminal is used to calculate the real-time entropy production rate of each subsystem based on the multi-dimensional physical field data of each subsystem in the target system collected in real time, and to generate an entropy production rate cloud map of each subsystem. The entropy production feedback module is used to determine the entropy production change rate and cumulative entropy production change rate of each subsystem based on the real-time entropy production rate of each subsystem, and to obtain the real-time control quantity of each subsystem based on the entropy production change rate and cumulative entropy production change rate of each subsystem. The flow topology optimization module is used to determine the high entropy production area in each subsystem based on the numerical distribution of the entropy production rate cloud map of each subsystem and the preset threshold rules, and to match the execution device responsible for regulating each high entropy production area according to the preset energy flow path and device topology relationship. The digital twin construction module is used to simulate and output the entropy production rate change trend of each subsystem under the action of the real-time control quantity, as well as the key performance parameters of the execution device responsible for controlling each high entropy production region in each subsystem, taking the real-time control quantity of each subsystem as input. The decision module is used to determine the actual control quantity of the execution device in each subsystem that matches the high-entropy-producing region, based on the real-time control quantity of each subsystem, the entropy production rate change trend of each subsystem, and the key performance parameters of the execution device responsible for controlling each high-entropy-producing region in each subsystem, on the premise of meeting the equipment safety threshold and / or system stability boundary of each subsystem, and to send the actual control quantity of the execution device to the corresponding execution mechanism, so as to minimize the real-time entropy production rate of each subsystem.

2. The system as described in claim 1, characterized in that, The decision-making module includes at least an actual control quantity determination unit, which is used for: Based on the entropy production rate change trend of each subsystem, determine the control coefficient of each subsystem; Based on the key performance parameters of the execution equipment responsible for regulating each high-entropy production region in each subsystem, determine the adjustable margin of the execution equipment in each high-entropy production region. Based on the premise of satisfying the equipment safety threshold and / or system stability boundary of each subsystem, the actual control amount of the execution equipment in each subsystem that matches the high-entropy production region is determined according to the real-time control amount of each subsystem, the control coefficient of each subsystem, and the adjustable margin of the execution equipment in each high-entropy production region.

3. The system as described in claim 2, characterized in that, The steps for determining the adjustable margin of the execution equipment in each high-entropy production region include: Based on the entropy production rate cloud map of each subsystem, determine the entropy production level corresponding to the high entropy production region in each subsystem; Based on the entropy production level corresponding to the high entropy production region in each subsystem, determine the maximum adjustable margin of the execution equipment in each high entropy production region; Based on the operating parameters of the execution devices in each high-entropy production region and the maximum adjustable margin of the execution devices in each high-entropy production region, and on the premise of satisfying the safety threshold of each execution device, the adjustable margin of the execution devices in each high-entropy production region is determined.

4. The system as described in claim 1, characterized in that, The entropy production feedback module includes at least an adaptive gain unit, which is used for: The entropy production rate and entropy production rate gain of each subsystem are determined by the derivative element in the adaptive gain PID control method. The cumulative entropy productivity and cumulative entropy productivity gain of each subsystem are determined by the integral element in the adaptive gain PID control method. The real-time control quantity of each subsystem is obtained based on the entropy production change rate, entropy production change rate gain, cumulative entropy production rate, and cumulative entropy production rate gain of each subsystem.

5. The system as described in claim 4, characterized in that, The The flow topology optimization module includes at least an execution device determination unit, which is used for: When the entropy production change rate of a certain subsystem is greater than 0, the acceleration fan responsible for regulating each high entropy production region in that subsystem will be used as the execution device for regulating each high entropy production region in that subsystem. If the cumulative entropy production rate of a certain subsystem exceeds a preset cumulative threshold, the cleaning equipment responsible for regulating each high-entropy production area in that subsystem will be used as the execution equipment for regulating each high-entropy production area in that subsystem.

6. The system of claim 5, wherein the adaptive gain unit is further configured to: The real-time load rate of each subsystem is determined based on the maximum rated power and real-time output power of each subsystem. The real-time load change rate of each subsystem is determined based on the real-time load rate of each subsystem. If the real-time load change rate of the first subsystem exceeds the preset threshold load change rate, the reward value of the first subsystem is generated based on the real-time entropy production rate, temperature gradient, vorticity and pressure difference of the first subsystem through the preset reward function. Based on the range of the reward value of the first subsystem, determine the correction direction and adjustment range of the entropy production change rate gain and the cumulative entropy production change rate gain of the first subsystem. Based on the correction direction and the adjustment magnitude, adjust the entropy production rate change gain and the cumulative entropy production rate change gain of the first subsystem to maximize the reward value of the first subsystem. The entropy change rate gain and cumulative entropy deviation gain corresponding to the maximum reward value of the first subsystem are determined as the entropy change rate gain and cumulative entropy deviation gain of the first subsystem.

7. A Digital Twin with Entropy Production Constraints The optimization method is characterized by, include: Based on the multi-dimensional physical field data of each subsystem in the target system collected in real time, the real-time entropy production rate of each subsystem is calculated, and an entropy production rate cloud map of each subsystem is generated. Based on the real-time entropy production rate of each subsystem, determine the entropy production change rate and cumulative entropy production change rate of each subsystem, and obtain the real-time control quantity of each subsystem based on the entropy production change rate and cumulative entropy production change rate of each subsystem. Based on the numerical distribution of the entropy productivity cloud map of each subsystem and the preset threshold rules, the high entropy production area of ​​each subsystem is determined, and the execution device responsible for regulating each high entropy production area is matched according to the preset energy flow path and device topology relationship. Using the real-time control quantity of each subsystem as input, the simulation outputs the entropy production rate change trend of each subsystem under the action of the real-time control quantity, as well as the key performance parameters of the execution equipment responsible for controlling each high entropy production region in each subsystem. Based on the real-time control amount of each subsystem, and through the entropy production rate change trend of each subsystem and the key performance parameters of the execution equipment responsible for controlling each high-entropy production region in each subsystem, and on the premise of meeting the equipment safety threshold and / or system stability boundary of each subsystem, the actual control amount of the execution equipment matching the high-entropy production region in each subsystem is determined, and the actual control amount of the execution equipment is sent to the corresponding execution mechanism to minimize the real-time entropy production rate of each subsystem.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the digital twin of the entropy production constraint as described in claim 7. Optimization methods.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the digital twin of the entropy production constraint as described in claim 7. Optimization methods.

10. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the digital twin with entropy production constraints as described in claim 7. Optimization methods.

Images