A heat storage management system and method based on edge computing and temperature difference information

Abstract

The application discloses a heat storage management system and method based on edge computing and temperature difference information, and relates to the technical field of edge computing; the system comprises a data sensing and collecting module, a trend analysis and identification module, a multi-stage heat exchange path decision module and an edge execution and feedback module; the technical points are as follows: on the one hand, the system is operated to identify the temperature difference evolution trend in real time through edge computing and dynamically switch the multi-stage heat exchange path, so that super-large temperature difference heat storage is realized, and the water tank volume is reduced under the same heat storage capacity; on the other hand, the edge intelligence is used to accurately control the ice-water mixture ratio in the near phase change area, so that the pipe freezing risk is avoided, the unit water body heat storage density is significantly improved, and the non-steady-state high-density heat storage effect that cannot be achieved by a conventional water heat storage system is realized.

Description

Technical Field

[0001] This invention relates to the field of edge computing technology, specifically to a thermal storage management system and method based on edge computing and temperature difference information. Background Technology

[0002] Edge computing is a computing model that pushes data processing capabilities down to local devices or network edge nodes closer to the data source. Its core lies in reducing the long-distance transmission of data to remote cloud centers, enabling low-latency, high-real-time localized analysis and decision-making. In scenarios such as industrial control and intelligent sensing, edge computing nodes can be directly deployed on the field equipment side to collect and process sensor data in real time and execute control commands, and can still operate independently even when the network is interrupted. Edge computing technologies are suitable for applications with high security requirements, such as temperature difference monitoring and multi-stage heat exchange control of thermal storage systems.

[0003] Traditional thermal storage systems generally rely on preset thresholds for start-up and shutdown control, lacking the ability to accurately perceive the dynamic evolution of the temperature field within the storage tank, resulting in lag in regulation. For example, when the water temperature in the upper layer of the storage tank slowly drops from 60°C to 58°C, the system cannot recognize that it is about to enter the mid-temperature range and continues to maintain the direct supply mode of the plate heat exchanger until the temperature drops below 55°C before passively starting the heat pump, causing fluctuations in the heating temperature. At the same time, the heat exchange path is fixed and cannot automatically switch between single-stage and dual-stage heat pumps according to changes in temperature zones, causing the heat pump to operate at low loads for a long time when extracting heat in the low-temperature range, resulting in a significant decrease in COP and an increase in energy consumption. In addition, the control strategy is crude, often using simple PID regulation, which is prone to large frequency oscillations when faced with flow disturbances or sudden load changes, affecting the equipment lifespan. Moreover, it lacks a graded response mechanism, and when the bottom water temperature suddenly drops below 3°C or the circulation is interrupted, it cannot promptly implement antifreeze protection, posing safety hazards such as ice blockage or dry burning. Summary of the Invention

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] A thermal storage management system based on edge computing and temperature difference information includes:

[0006] The data sensing and acquisition module is used to acquire operating parameters, including the water temperature of the upper, middle and lower layers of the thermal storage tank, which are preprocessed to form a basic dataset.

[0007] The trend analysis and identification module is used to construct a temperature gradient evolution model inside the heat storage tank based on the basic dataset, identify the temperature range to which the current heat storage state belongs, and output the evolution inference results including the expected time to enter the next temperature range.

[0008] The multi-stage heat exchange path decision module is used to dynamically select the optimal heat exchange path among plate heat exchanger direct supply, single-stage heat pump or double-stage heat pump based on the evolution reasoning results, combined with user-side water supply requirements and electricity price time-of-use signals, so as to generate path switching instructions.

[0009] The edge execution and feedback module is used to receive the path switching command, drive the actuators including the electric regulating valve and the frequency converter pump to move, and monitor the deviation between the execution result and the target requirement in real time. When the deviation exceeds the limit, a secondary adjustment mechanism is triggered.

[0010] A temperature zone synchronization control signal generator is used to generate a periodic control signal whose period is associated with the expected time to enter the next temperature zone;

[0011] The edge execution and feedback module is further configured to: during the transition from the current temperature zone to the next temperature zone, perform cyclic driving and state sampling of the actuator in the currently activated heat exchange path according to the control signal, and change the average output power of the actuator by adjusting the pulse width or frequency of the control signal, so as to achieve precise stabilization of water supply temperature or flow rate.

[0012] Furthermore, the control signal is a pulse density modulation signal, the pulse density of which is proportional to the required actuator action intensity, used to achieve gradual adjustment of the opening of the electric regulating valve or smooth control of the variable frequency pump speed.

[0013] Furthermore, each cycle of the control signal is divided into an energy injection period, a system response period, and a state sampling period;

[0014] During the energy injection period, full drive power is output to the actuator to rapidly change its state;

[0015] During the system response period, the driving power is stopped or reduced, and the thermodynamic system is allowed to reach a new equilibrium under its own inertia.

[0016] During the state sampling period, key parameters are collected to evaluate the control effect and determine the signal parameters for the next cycle.

[0017] The duration of the system response period is dynamically set based on the thermal inertia of the water in the thermal storage tank and the type of heat exchange path currently in operation, to ensure that the system has stabilized before the state sampling.

[0018] The system is configured to: calculate the instantaneous heat transfer efficiency index during the state sampling period, and dynamically adjust the duration of the energy injection period in the next cycle based on the deviation of the index from the expected value, so as to achieve closed-loop optimization of the heat transfer process.

[0019] Furthermore, the generation period of the temperature zone synchronization control signal is determined by the expected time to enter the next temperature zone and the thickness of the current temperature zone; the shorter the expected transition time or the smaller the temperature zone thickness, the higher the frequency of the signal and the finer the control granularity.

[0020] Furthermore, the drive circuit of the actuator integrates a current feedback unit; the edge execution and feedback module indirectly determines whether the valve is stuck or the pump is running dry by monitoring changes in the drive current, and uses this information as a feedback signal to adjust the control signal.

[0021] Furthermore, the operating parameters of the temperature zone synchronous control signal generator are modulated by the phase change early warning signal output by the trend analysis and identification module; when a phase change early warning occurs, the signal frequency is automatically increased and the intensity of a single energy injection is reduced, so as to achieve precise control of anti-freezing and blockage in the near phase change zone.

[0022] Furthermore, the multi-stage heat exchange path decision module generates a corresponding initial control strategy while outputting the path switching command; the temperature zone synchronous control signal generator initializes the frequency and pulse width of the signal based on this strategy.

[0023] Furthermore, the secondary adjustment mechanism works in conjunction with the temperature zone synchronous control signal:

[0024] First-level regulation involves real-time compensation through fine-tuning the control signal parameters of the current cycle;

[0025] Secondary regulation: When primary regulation fails and the water temperature is about to cross the temperature zone boundary, the trend analysis and identification module is triggered to make a new prediction and request the multi-stage heat exchange path decision module to make a new decision.

[0026] The three-stage regulation system, when it detects that the lower water temperature has entered the near-phase change zone and continues to decline, forcibly overrides the current regulation signal and executes an emergency protection action to shut down the valve.

[0027] The present invention also provides a thermal storage management method based on edge computing and temperature difference information, comprising the following steps:

[0028] The operating parameters, including the water temperatures of the upper, middle, and lower layers of the thermal storage tank, are obtained and preprocessed to form a basic dataset.

[0029] Based on the aforementioned basic dataset, a temperature gradient evolution model inside the thermal storage tank is constructed to identify the temperature range to which the current thermal storage state belongs and to predict the evolutionary inference results.

[0030] Based on the evolutionary reasoning results, combined with the user-side water supply requirements and time-of-use electricity price signals, the optimal heat exchange path and heat pump operating stage are dynamically selected to generate path switching instructions;

[0031] Upon receiving the path switching instruction, the pre-configured actuator is driven to perform the action, and the deviation between the execution result and the target requirement is monitored in real time. When the deviation is outside the corresponding standard threshold, a secondary adjustment mechanism is triggered.

[0032] This also includes:

[0033] A periodic control signal is generated, the period of which is synchronized with the temperature transition time predicted based on temperature difference information;

[0034] During the temperature transition, the actuators in the currently activated heat exchange path are cyclically driven and their status is sampled based on this control signal.

[0035] By adjusting the pulse width or frequency of the control signal, precise stabilization of water supply temperature or flow rate can be achieved.

[0036] Furthermore, a pulse density modulation signal is used as the control signal, and the average output power of the actuator is linearly controlled by changing the number of pulses per unit time;

[0037] Each cycle of the control signal is divided into an energy injection period, a system response period, and a state sampling period to match the driving force, inertial response, and evaluation program of the thermodynamic system, thereby achieving stable control.

[0038] During the state sampling period of each control signal cycle, an instantaneous energy efficiency coefficient is calculated based on the collected temperature and flow data and compared with the target energy efficiency coefficient. Based on this, the intensity of the energy injection period in the next cycle is adjusted to achieve dynamic energy efficiency optimization.

[0039] The phase change warning level is mapped to a dual constraint on the frequency and pulse width of the control signal. Under high risk, a high-frequency, low-pulse-width control mode is adopted to balance heat extraction efficiency and antifreeze safety.

[0040] The present invention has the following beneficial effects:

[0041] This solution constructs a one-dimensional unsteady-state heat transfer differential equation based on the temperature difference rate and heat transfer coefficient between the upper and lower layers, enabling accurate modeling of the temperature gradient evolution process within the thermal storage tank and improving the system's perception accuracy of thermal energy status. Simultaneously, the model incorporates water heat capacity, thermal diffusivity, and boundary conditions for dynamic prediction, allowing the system to issue early warnings before the water temperature enters the next temperature zone and initiate heat exchange path switching preparations in advance. This solves the problems of traditional water-based thermal storage systems that rely on empirical thresholds and lack dynamic prediction capabilities, leading to control lag and large energy efficiency fluctuations. It provides a highly reliable data foundation for subsequent intelligent decision-making and closed-loop regulation.

[0042] This solution is based on a multi-stage heat exchange path dynamic decision-making mechanism, which can automatically switch between plate heat exchanger direct supply, single-stage heat pump, two-stage heat pump and even subcooled water generation mode according to the heat storage status, so as to match the heat demand under different temperature zones to the maximum extent and improve the heat pump energy efficiency ratio (COP). On the other hand, the decision-making process introduces a fuzzy rule engine to achieve optimal control of operating costs, solves the problems of heat extraction difficulties, frequent start-stop of heat pumps and low energy utilization in traditional systems at low temperatures, and provides instruction basis for precise adjustment of the subsequent execution layer.

[0043] This scheme employs a three-level response secondary adjustment mechanism to implement tiered intervention when operational deviations occur: Level 1 adjustment performs proportional compensation based on the deviation amount to maintain stable system operation; Level 2 adjustment reactivates the trend identification and path decision module during sudden changes in operating conditions to achieve dynamic reconstruction of the control strategy; Level 3 adjustment executes emergency protection actions in extreme situations to prevent safety accidents such as ice blockage and dry burning. Simultaneously, this mechanism achieves quantitative adjustment, avoiding overshoot and oscillation problems common in traditional PID control, thus improving adjustment accuracy and system robustness. Furthermore, the operation of this scheme directly relies on the state identification results and decision commands provided in the two effect descriptions given above. Attached Figure Description

[0044] Figure 1 This is a schematic diagram of the actuator between the waste heat source and the heat user in this invention;

[0045] Figure 2 This is a schematic diagram comparing the operating states of the heating mode and the heat storage mode in this invention.

[0046] Figure 3 This is a schematic diagram comparing the operating states of plate heat exchanger, single-stage heat pump, and two-stage heat pump in this invention. Detailed Implementation

[0047] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0048] A thermal storage management system based on edge computing and temperature difference information includes: a data sensing and acquisition module for acquiring operating parameters including the water temperatures of the upper, middle, and lower layers of the thermal storage tank, which are preprocessed to form a basic dataset; a trend analysis and identification module for constructing a temperature gradient evolution model inside the thermal storage tank based on the basic dataset, identifying the temperature range to which the current thermal storage state belongs, and outputting an evolutionary inference result including the expected time to enter the next temperature range; a multi-stage heat exchange path decision module for dynamically selecting the optimal heat exchange path among plate heat exchanger direct supply, single-stage heat pump, or two-stage heat pump based on the evolutionary inference result, combined with user-side water supply requirements and time-of-use electricity price signals, to generate a path switching command; and an edge execution and feedback module. The module is used to receive the path switching command, drive the actuators including the electric regulating valve and the variable frequency pump, and monitor the deviation between the execution result and the target requirement in real time. When the deviation exceeds the limit, a secondary adjustment mechanism is triggered. The temperature zone synchronous control signal generator is used to generate a periodic control signal whose period is associated with the expected time to enter the next temperature zone. The edge execution and feedback module is further configured to: during the transition from the current temperature zone to the next temperature zone, perform cyclic driving and state sampling of the actuators in the currently activated heat exchange path according to the control signal, and change the average output power of the actuators by adjusting the pulse width or frequency of the control signal, so as to achieve precise stabilization of the water supply temperature or flow rate.

[0049] In the above embodiments, the data sensing and acquisition module functions to acquire operating parameters such as the water temperature of the upper, middle, and lower layers of the thermal storage tank, forming a basic dataset. Operating parameters extend far beyond water temperature, including tank pressure, inlet and outlet water flow rates, ambient temperature, user-side heat load, and a set of heat pump operating status parameters (such as compressor frequency and evaporator / condenser inlet and outlet temperatures). This data is collected through various locally deployed sensors (such as water temperature sensors, pressure sensors, and flow meters) and connected to an edge computing gateway via an industrial bus. The key to this data sensing and acquisition module lies in its "edge" attribute: data is preprocessed locally and has a caching mechanism, allowing it to operate independently even during network outages, ensuring the system's real-time performance and reliability. This provides an accurate, consistent, and low-latency data foundation for all subsequent advanced analyses.

[0050] The trend analysis and identification module is used to gain insights into dynamic trends from static data. Its core task is to construct a temperature gradient evolution model inside the thermal storage tank, identify the current temperature zone, and output the evolutionary inference results. It is a one-dimensional unsteady-state heat transfer differential equation, which comprehensively considers the rate of temperature difference change between the upper and lower water layers, the heat capacity characteristics of the water, and the effective heat transfer coefficient characterizing the overall heat exchange capacity within the tank. By solving this model, the system can not only calculate the current stored heat and remaining releasable heat, but also predict the "expected time to enter the next temperature zone" and perform a safety margin assessment (e.g., initiating a phase change warning when the lower water temperature approaches 10°C). This function upgrades the system's control logic from the traditional, lagging "threshold triggering" to proactive "trend prediction," providing a scientific basis for dynamic decision-making.

[0051] The multi-stage heat exchange path decision module transforms trend analysis results into actionable strategies. Based on evolutionary reasoning, combined with user-side water supply requirements and time-of-use electricity pricing signals, it dynamically selects the optimal heat exchange path. Its decision logic is a multi-objective optimization process: in the high-temperature range with low user demand, it selects the most energy-efficient plate heat exchanger for direct supply; in the medium-temperature range with higher user demand, it switches to a single-stage heat pump; in the low-temperature range, to maintain energy efficiency, it activates a two-stage heat pump; and in the near-phase change range, it innovatively activates the subcooling water generation function of the two-stage heat pump to extract latent heat. This decision process is not a simple table lookup; the manual indicates that it uses a fuzzy rule engine to comprehensively process multi-dimensional information such as tank temperature, load, cost, and energy efficiency, outputting the final valve and heat pump control commands, achieving a balance between technical and economic efficiency.

[0052] The edge execution and feedback module is used to put decisions into practice and supervise their execution. It receives path switching commands and drives actuators such as electric regulating valves and variable frequency pumps. These actuators are all integrated with safety interlock functions, such as travel limits and torque protection. More importantly, it has a feedback function: the module monitors the actual parameters (such as water temperature and flow rate) after execution in real time and compares them with the target requirements. To this end, the system sets a series of quantified standard thresholds (such as water temperature deviation ±2℃, flow rate deviation ±5%). Once the deviation exceeds the limit, a secondary adjustment mechanism is triggered. This ensures that control commands are not only executed, but also executed effectively, forming the first layer of protection for the system's closed-loop control.

[0053] A temperature zone synchronous control signal generator and its coordinated control are used to generate a periodic control signal, the period of which is correlated with the "expected time to enter the next temperature zone". This means that the control granularity of the system is adaptive to the dynamic changes of the thermodynamic process—the shorter the transition time, the more frequent the control. The edge execution and feedback module is further configured to cyclically drive and sample the actuators based on this control signal during the temperature zone transition. Specifically, by adjusting the pulse width or frequency of the signal, the average output power of the actuators is changed, thereby ultimately achieving precise stabilization of the water supply temperature or flow rate. This control method fully considers the response characteristics of a large inertial system, and through an intermittent "drive-wait-observe" cycle synchronized with the system's thermal dynamics, it avoids the system oscillations and actuator wear that are easily caused by traditional continuous control, achieving smooth and precise regulation.

[0054] Furthermore, the control signal is a pulse density modulation signal, the pulse density of which is proportional to the required actuator action intensity, used to achieve gradual adjustment of the opening of the electric regulating valve or smooth control of the variable frequency pump speed.

[0055] In the above embodiments, the control signal is the core control command driving actuators such as electric regulating valves and variable frequency pumps. Pulse density modulation (PDM) is a technique that adjusts average power by changing the number of pulses per unit time. Unlike common pulse width modulation (PWM), which adjusts the duty cycle by changing the width of a single pulse, PDM uses a fixed pulse width and transmits control quantities through its distribution density on the time axis, enabling gradual and smooth control of actuators.

[0056] For electrically operated regulating valves, traditional on / off or simple proportional control can cause the valve to open or close abruptly, resulting in drastic fluctuations in water flow and pressure, and consequently, shocks in the water supply temperature. Using PDM signals, the system can drive the valve motor to open in small, multiple steps by outputting a dense series of pulses, or to slowly close it by outputting sparse pulses. For variable frequency pumps, PDM signals can serve as a high-level control input to the inverter, linearly setting the average operating speed of the motor through pulse density. The denser the pulses, the higher the commanded average speed, and the smoother the pump's flow rate and head; conversely, a denser pulse results in a smoother decrease.

[0057] Furthermore, each cycle of the control signal is divided into an energy injection period, a system response period, and a state sampling period;

[0058] During the energy injection period, full drive power is output to the actuator to rapidly change its state;

[0059] During the system response period, the driving power is stopped or reduced, and the thermodynamic system is allowed to reach a new equilibrium under its own inertia.

[0060] During the state sampling period, key parameters are collected to evaluate the control effect and determine the signal parameters for the next cycle.

[0061] The duration of the system response period is dynamically set based on the thermal inertia of the water in the thermal storage tank and the type of heat exchange path currently in operation, to ensure that the system has stabilized before the state sampling.

[0062] The system is configured to: calculate the instantaneous heat transfer efficiency index during the state sampling period, and dynamically adjust the duration of the energy injection period in the next cycle based on the deviation of the index from the expected value, so as to achieve closed-loop optimization of the heat transfer process.

[0063] In the above embodiments, the energy injection period: During this stage, the system outputs full drive power to the actuator (such as an electric valve or a variable frequency pump). The purpose is to overcome the static friction or inertia of the system to quickly change the state of the actuator, for example, to start the valve turning or to accelerate the pump. This is a stage where active control is applied.

[0064] System Response Period: After energy injection, the system stops or significantly reduces its drive power. This is to allow the thermal system to reach a new equilibrium due to its own inertia. The water, pipes, and heat exchangers within the heat storage tank have enormous heat capacity, and changes in temperature and flow rate require time to propagate. Continuous drive would result in transient, erratic values ​​measured by the sensors, easily leading to misjudgments by the control system and causing overshoot or oscillations. This "waiting period" provides the thermal system with the necessary response time. The duration of this response period is dynamically set based on the thermal inertia of the water in the heat storage tank and the type of heat exchange path currently in operation, ensuring that the system has stabilized before entering state sampling.

[0065] State sampling period: After the system stabilizes, key operating parameters (such as actual water supply temperature and flow rate) are collected during this stage to evaluate the control effect. More importantly, the system calculates an instantaneous heat transfer efficiency index (such as the coefficient of performance) during this stage, and uses the deviation of this index from the expected value as the basis for deciding the signal parameters for the next cycle. For example, if the actual temperature is still lower than the target, the duration of the energy injection period is increased in the next cycle; if the target has been reached, the injection time is maintained or decreased.

[0066] Furthermore, the generation period of the temperature zone synchronization control signal is determined by the expected time to enter the next temperature zone and the thickness of the current temperature zone; the shorter the expected transition time or the smaller the temperature zone thickness, the higher the frequency of the signal and the finer the control granularity.

[0067] In the above embodiments, the specific determination of the temperature zone synchronization control signal period is based on the expected time of entering the next temperature zone and the thickness of the current temperature zone. Synchronizing with the "expected time of entering the next temperature zone," for example, when the model predicts that the water temperature will transition from one temperature zone to another in 30 minutes, the system may fine-tune at a relatively low frequency (e.g., one cycle every 5 minutes) to smoothly guide this transition. Conversely, if the prediction indicates that the transition will occur within 5 minutes, the system will automatically increase the frequency of the control signal significantly (e.g., one cycle every 30 seconds) to respond to rapidly changing operating conditions with more intensive control actions, preventing the water temperature from running out of control and crossing the boundary.

[0068] Considering the current temperature range, a control buffer zone is established. A higher temperature range (e.g., the high-temperature section starting above 60°C) provides a larger buffer space, allowing the system to adjust with coarser control granularity. Conversely, a lower temperature range (e.g., the near-phase transition section at only 0-10°C), especially near the phase transition point, has extremely small tolerance and requires very fine control. In this case, even with the same transition time, the system will automatically adopt a higher control frequency due to the smaller temperature range. This mechanism of dynamically adjusting control granularity ensures that the system always operates with fine precision. It avoids unnecessary frequent adjustments under steady-state or slowly changing conditions, while guaranteeing sufficiently strong control capabilities to ensure system stability and safety during drastic changes in operating conditions or in critical sensitive areas.

[0069] Furthermore, the drive circuit of the actuator integrates a current feedback unit; the edge execution and feedback module indirectly determines whether the valve is stuck or the pump is running dry by monitoring changes in the drive current, and uses this information as a feedback signal to adjust the control signal.

[0070] In the above embodiments, the drive current of the electric actuator and pump is within a expected range during normal operation. When the valve fails to operate due to mechanical jamming or blockage by foreign objects, the drive motor will experience an abnormally high current due to stalling; conversely, when the pump runs dry (e.g., air entering the pipeline causes a sharp decrease in load), its operating current will be significantly lower than normal. By monitoring the current in real time and comparing it with the normal range, the system can achieve early, online diagnosis of these common faults without the need for additional physical sensors.

[0071] Once the system identifies an abnormal state through current feedback, it doesn't simply issue an alarm; instead, it proactively uses this information as a feedback signal to adjust the control signals. For example, if a slight sign of valve jamming is detected, the system might temporarily increase the power or duration of the energy injection period in the next control cycle, attempting to "overcome" the jamming point with greater torque. If it determines that the pump is running dry, the system might immediately adjust the signal, reducing the pump speed or triggering higher-level protection logic to prevent equipment damage.

[0072] Furthermore, the operating parameters of the temperature zone synchronous control signal generator are modulated by the phase change early warning signal output by the trend analysis and identification module; when a phase change early warning occurs, the signal frequency is automatically increased and the intensity of a single energy injection is reduced, so as to achieve precise control of anti-freezing and blockage in the near phase change zone.

[0073] In the above embodiments, increasing the signal frequency means that in high-risk areas close to 0°C, the system will shorten the control cycle and perform more frequent "observation-decision-action" loops. This allows the system to monitor minute changes in water temperature more closely and react quickly to any signs of freezing.

[0074] While increasing the monitoring frequency, the system reduced the intensity of each control action. For example, when driving a valve, the rotation was changed from 10 degrees to only 2 degrees; when controlling the pump speed, only 1 Hz was adjusted instead of 5 Hz. This "small-step fine-tuning" mode aims to avoid the water becoming too cold due to excessively strong single control actions, which could instantly form a large number of ice crystals, thus greatly reducing the risk of pipe freezing and blockage.

[0075] Furthermore, the multi-stage heat exchange path decision module generates a corresponding initial control strategy while outputting the path switching command; the temperature zone synchronous control signal generator initializes the frequency and pulse width of the signal based on this strategy.

[0076] In the above embodiments, a collaborative initialization mechanism exists between the multi-stage heat exchange path decision module and the temperature zone synchronous control signal generator. Specifically, the multi-stage heat exchange path decision module generates a corresponding initial control strategy while outputting the path switching command; then, the temperature zone synchronous control signal generator initializes the signal frequency and pulse width based on this strategy. This design ensures that the actuator control can follow up instantly and smoothly when the system switches heat exchange paths. Each path switch (such as switching from direct plate heat exchanger supply to a single-stage heat pump) means that the system's operating point, thermal inertia, and dynamic characteristics change. Different paths have different optimal control parameters (such as signal frequency and pulse width). For example, when switching to a slower-responding heat pump mode, the initial strategy may suggest a lower control frequency and a longer system response time; while when switching to a faster-responding direct plate heat exchanger mode, the initial strategy may set a higher control frequency. The initial control strategy is pre-set based on the known characteristics of different heat exchange paths.

[0077] Furthermore, the secondary adjustment mechanism works in conjunction with the temperature zone synchronous control signal: primary adjustment, which performs real-time compensation by fine-tuning the control signal parameters of the current cycle; secondary adjustment, which triggers the trend analysis and identification module to re-predict when the primary adjustment is ineffective and the water temperature is about to cross the temperature zone boundary, and requests the multi-stage heat exchange path decision module to re-determine; tertiary adjustment, which forcibly overrides the current control signal and executes the emergency protection action of shutting down the valve when the lower water temperature is detected to enter the near-phase change section and continue to decline.

[0078] In the above embodiments, primary regulation provides real-time compensation by fine-tuning the control signal parameters (such as pulse width or frequency) of the current cycle. For example, when the water supply temperature is detected to be slightly lower than the target, the pulse density driving the valve is slightly increased in the next control cycle to slightly increase the flow rate. This primary regulation is designed to handle small, normalized deviations and disturbances and is the main means for the system to maintain daily stability.

[0079] When primary regulation fails, and the deviation persists or even worsens, causing the water temperature to approach the temperature zone boundary, secondary regulation is triggered. At this point, the system triggers the trend analysis and identification module to perform a re-prediction and requests the multi-stage heat exchange path decision module to re-determine. It determines whether the current heat exchange path is no longer applicable and whether a switch to another path is necessary (e.g., from a single-stage heat pump to a two-stage heat pump). This is a significant adjustment by the system to address substantial changes in operating conditions.

[0080] Level 3 regulation is the highest level of safety protection. This level is activated when the lower water temperature is detected to be entering the near-phase change zone and continuing to drop, indicating an imminent risk of freezing. At this point, the system will forcibly override all current control signals and execute emergency protection actions such as shutting down valves. This is a hard protection mechanism with the highest priority.

[0081] This invention also provides a thermal storage management method based on edge computing and temperature difference information, comprising the following steps: acquiring operating parameters including the water temperatures of the upper, middle, and lower layers of the thermal storage tank, and preprocessing them to form a basic dataset; based on the basic dataset, constructing a temperature gradient evolution model inside the thermal storage tank, identifying the temperature zone to which the current thermal storage state belongs, and predicting the evolution inference result; based on the evolution inference result, combined with the user-side water supply requirements and time-of-use electricity price signals, dynamically selecting the optimal heat exchange path and heat pump operating stage to generate a path switching command; receiving the path switching command, driving the pre-configured actuator to act, and monitoring the deviation between the execution result and the target requirement in real time; when the deviation is outside the corresponding standard threshold, triggering a secondary adjustment mechanism; further comprising: generating a periodic control signal, the period of which is synchronized with the temperature zone transition time predicted based on temperature difference information; during the temperature zone transition, according to this control signal, cyclically driving and sampling the actuator in the currently activated heat exchange path; and achieving precise stabilization of water supply temperature or flow rate by adjusting the pulse width or frequency of the control signal.

[0082] Furthermore, a pulse density modulation signal is used as the control signal, and the average output power of the actuator is linearly controlled by changing the number of pulses per unit time. Each cycle of the control signal is divided into an energy injection period, a system response period, and a state sampling period to match the drive, inertial response, and evaluation program of the thermodynamic system, thereby achieving stable control. During the state sampling period of each control signal cycle, an instantaneous energy efficiency coefficient is calculated based on the collected temperature and flow data and compared with the target energy efficiency coefficient. Based on this, the intensity of the energy injection period of the next cycle is adjusted to achieve dynamic energy efficiency optimization. The phase change warning level is mapped to a dual constraint on the frequency and pulse width of the control signal. Under high risk, a high-frequency, small-pulse-width control mode is adopted to balance heat extraction efficiency and antifreeze safety.

[0083] Through a trend analysis and identification module, a one-dimensional unsteady-state heat transfer differential equation inside the thermal storage tank is constructed and solved. This model integrates real-time water temperature data from both upper and lower layers, water heat capacity, and heat transfer coefficients obtained through on-site calibration, enabling dynamic simulation of the temperature field's evolution in the vertical direction. Unlike traditional systems that rely solely on instantaneous temperature values, this trend analysis and identification module not only accurately identifies the current temperature zone (high temperature, medium temperature, low temperature, near-phase transition zone) of the thermal storage state but also predicts the expected time to enter the next temperature zone. This predictive mechanism represents an upgrade from feedforward control to proactive control. The system's operating logic shifts from a passive response of "the water temperature has dropped to 55°C, therefore the heat pump must be started" to an active planning of "the model predicts the water temperature will enter the medium temperature zone in 15 minutes, therefore the heat pump sequence should be started now." This provides a valuable preparation window for subsequent multi-stage heat exchange path switching, allowing for a smooth and orderly transition and completely avoiding heating temperature fluctuations or brief interruptions caused by control lag. For example, when transitioning from a high-temperature range to a medium-temperature range, the system can preheat the heat pump unit and preset valve openings to ensure a seamless switch at the critical point, with virtually no impact on the user. Furthermore, this predictive capability is deeply integrated with economic operations; the system can more accurately plan heat storage strategies during periods of low electricity prices based on predicted temperature zone dwell time.

[0084] When the water temperature is in the high-temperature range and the user's temperature demand is not high, the most energy-efficient plate heat exchanger is prioritized for direct supply. When the water temperature drops to the medium-temperature range and the user's demand is high, the system switches to a single-stage heat pump for heating. When entering the low-temperature range, in order to maintain the overall system energy efficiency, a two-stage heat pump is activated to increase the compression ratio. In the near-phase change range, the system decides to activate the low-temperature stage evaporator subcooling water generation function of the two-stage heat pump, breaking through the limits of traditional sensible heat utilization and extracting some latent heat while strictly controlling the ice-water mixture ratio (≤10%). This temperature-dependent decision-making mechanism ensures that the main energy-consuming equipment, such as the heat pump, always operates in the high-efficiency range, avoiding a sharp drop in COP and a surge in energy consumption caused by forced operation in the low-temperature range.

[0085] The edge execution and feedback module incorporates a three-level responsive secondary adjustment mechanism, defining a clear, progressively escalating control loop: when a deviation between the execution result and the target requirement is detected, first-level regulation is triggered, finely adjusting the compressor frequency or pump speed according to a preset proportional coefficient to attempt to quickly eliminate small-scale disturbances. If the deviation persists or worsens, indicating a potential significant change in operating conditions, second-level regulation is initiated. The core feature of this level is the re-invocation of the upstream trend analysis and identification module and the multi-level heat exchange path decision module, essentially reassessing the system state and making strategic-level re-decision, such as switching heat exchange paths. This empowers the system with the ability to dynamically reconfigure control strategies to cope with unexpected situations such as sudden load changes and equipment performance degradation. Finally, when extreme risks such as the lower-level water temperature entering the near-phase change zone and continuously decreasing are detected, third-level regulation immediately activates, executing mandatory protective actions (such as shutting down valves). This technology, by establishing a progressive response system from "fine-tuning compensation" to "strategy reconstruction" and then to "hard protection for safety," not only effectively avoids the oscillation and overshoot problems that are easily generated by traditional single PID control in complex thermal systems, but also builds a deep defense system, which greatly improves the stability of the system when facing internal and external interference and the safety when facing extreme risks.

[0086] This invention, by precisely controlling the proportion of ice crystal formation, safely utilizes the large amount of latent heat (approximately 334 kJ / kg) released by water during phase change, ensuring that continuous ice sheets do not form and block the pipes. This allows the system to push the lower limit of effective heat extraction close to 0°C, achieving ultra-large temperature difference heat storage and release from 95°C to near 0°C. When storing the same amount of heat, the required water tank volume can be significantly reduced (more than 45% as mentioned in the embodiments), greatly reducing initial equipment investment and floor space. Simultaneously, this also provides stronger heating guarantee capabilities for the system in extremely cold environments or during the final heat extraction phase, transforming previously considered "waste heat" low-temperature water into a usable high-density energy storage medium.

[0087] The present invention will now be described in detail with reference to the accompanying drawings.

[0088] Example 1:

[0089] Please see Figures 1 to 3 This embodiment provides a thermal storage management system based on edge computing and temperature difference information. The system deploys edge computing nodes locally in the thermal storage tank to collect multi-dimensional data such as thermal storage body temperature, phase change state, flow rate, pressure and external load demand in real time. Combined with a preset multi-level heat exchange strategy model, it dynamically decides the heat pump operation mode and valve switching logic to achieve autonomous optimization control of the entire process under thermal storage / heating conditions.

[0090] The system specifically includes the following sequentially operating functional modules, with explanations provided:

[0091] I. Data Sensing and Acquisition Module:

[0092] The operating parameters of the target thermal storage system are obtained and preprocessed to form a basic dataset.

[0093] Operating parameters include: the water temperatures of the upper, middle, and lower layers of the target thermal storage tank; tank pressure; inlet and outlet water flow rates; ambient temperature; user-side heat load; and a set of heat pump operating status parameters. These also include conventional parameters such as supply water temperature. The sources of these operating parameters are explained as follows: the water temperatures of the upper, middle, and lower layers of the target thermal storage tank are obtained through pre-configured water temperature sensors; the tank pressure is obtained through a tank pressure sensor; and subsequent parameters are obtained sequentially from: inlet and outlet water flow meter signals, ambient temperature sensor signals, user-side heat load demand signals, and the heat pump system. The heat pump operating status parameter set includes: compressor frequency and evaporator / condenser inlet and outlet temperatures. It should be noted that the aforementioned sensors... The sampling period is set according to actual needs and connected to a local edge computing gateway via an industrial bus. The edge computing gateway has a built-in data caching mechanism, which can store historical data for a preset duration locally when communication is interrupted, and automatically re-upload it to the cloud when the network is restored. The preset duration is usually set to 48 hours or more. Preprocessing of operating parameters includes outlier filtering, unit conversion, and timestamp alignment to ensure the accuracy and consistency of subsequent input data, thereby forming the required basic dataset. For example, at 08:00 on a winter heating day, the system detects that the water temperature of the target thermal storage tank is 92.3℃ in the upper layer, 89.1℃ in the middle layer, and 86.7℃ in the lower layer, with a flow rate of 120m³. 3 The user's heat load is 1.8MW and the ambient temperature is -8℃. These operating parameters are verified by the edge gateway and then packaged and uploaded for easy access to the next module.

[0094] II. Trend Analysis and Identification Module:

[0095] Based on the basic dataset, a temperature gradient evolution model is constructed inside the target thermal storage system to identify the current thermal storage state and predict the evolution inference results.

[0096] Specifically, the process of running the temperature gradient evolution model is as follows:

[0097] By calculating the temperature difference rate between the upper and lower layers and combining the water's heat capacity and heat transfer coefficient, a one-dimensional unsteady-state heat transfer differential equation is established to solve for the current stored heat Q and the remaining releaseable heat ΔQ. remaining The heat capacity of water refers to the amount of heat absorbed (or released) per unit mass of water when the temperature increases (or decreases) by 1°C, denoted as c. pThe unit is J / (kg·℃); within the normal pressure range, the specific heat capacity of water is approximately constant; in this embodiment, the water heat capacity is used to convert the mass and temperature change of the water body with its heat content; the mass m of any water layer in the heat storage tank within the target heat storage system is determined by the volume V and density ρ, and the calculation basis for its heat storage Q is: ; where m i Let T be the mass of the i-th water layer. i T represents the water temperature of the i-th layer. ref For reference temperature, i = 1, 2, ..., n. In this embodiment, there are three water layers, so n = 3. The above calculations are used to accurately solve for the heat storage capacity of the layers.

[0098] The heat transfer coefficient refers to the effective combined conductive and convection heat transfer coefficient between the water bodies inside the thermal storage tank, denoted as h. eff The unit is W / (m 2 (×℃); Due to natural convection, thermal stratification, and local disturbances in the water within the thermal storage tank, its heat transfer behavior is not pure conduction, but rather a coupled process of conduction and convection; therefore, a heat transfer coefficient is introduced to characterize the comprehensive heat exchange capacity between the upper and lower water layers; this coefficient is obtained through on-site calibration: under known heat input power P, water volume V, and initial temperature difference ΔT0, the temperature difference decay rate is monitored, and h is derived from this. eff The specific range is 50~200; its specific value varies with the intensity of water disturbance and the tank structure (such as whether or not a baffle is installed);

[0099] The one-dimensional unsteady heat transfer differential equation serves as the core of the temperature field evolution model. Assuming that the temperature inside the heat storage tank changes only along the vertical direction and neglecting the radial and circumferential gradients, the established partial differential equation is as follows: In the formula, T(z, t) represents the water temperature at position z and time t; α = k / ρ × c p : represents the thermal diffusivity of water, and k is the thermal conductivity of water; the equation is solved discretly using the finite difference method to predict the temperature distribution, and then the following can be calculated: Among them, T min H represents the minimum allowable heat extraction temperature of the system, where H is the height.

[0100] The criteria for identifying the current heat storage state range are as follows: heat storage tank water temperature ≥ 60℃ is marked as high temperature range, heat storage tank water temperature in the range of [40℃, 60℃) is marked as medium temperature range, heat storage tank water temperature in the range of [10℃, 40℃) is marked as low temperature range, and heat storage tank water temperature in the range of [0℃, 10℃) is marked as near-phase change range;

[0101] The specific output of the predicted evolutionary inference results is as follows:

[0102] The current temperature zone is marked, the expected time to enter the next temperature zone is set, and the safety margin assessment results are provided. The safety margin assessment results are based on the following process: when the water temperature in the lower layer of the target heat storage tank is detected to be approaching 10°C and continuing to decrease, the phase change early warning mechanism is activated to determine whether there is a tendency for local freezing. Conversely, under some ideal conditions, it indicates that the safety margin is sufficient. Other safety margin assessments will not be elaborated on here.

[0103] Results: By constructing a one-dimensional unsteady-state heat transfer differential equation based on the temperature difference rate and heat transfer coefficient between the upper and lower layers, the system accurately models the temperature gradient evolution process within the thermal storage tank, enabling real-time solutions and significantly improving the system's accuracy in sensing the thermal energy state. Simultaneously, the model incorporates water heat capacity, thermal diffusivity, and boundary conditions for dynamic prediction, allowing the system to issue early warnings before the water temperature enters the next temperature zone and initiate heat exchange path switching preparations in advance. This avoids heating interruptions due to response lag, thus solving the problems of traditional water thermal storage systems that rely on empirical thresholds and lack dynamic prediction capabilities, leading to control lag and large energy efficiency fluctuations. It provides a highly reliable data foundation for subsequent intelligent decision-making and closed-loop regulation, serving as a prerequisite for the coordinated operation of the entire system.

[0104] III. Multi-stage heat exchange path decision module:

[0105] Based on the output evolutionary reasoning results, combined with the external waste heat source temperature, user-side water supply requirements, and time-of-use electricity price signals, the optimal heat exchange path and heat pump operating stage are dynamically selected to generate control commands; the specific decision-making logic is as follows:

[0106] When the water temperature in the heat storage tank is in the high-temperature range and the user's water supply temperature is ≤55℃, the plate heat exchanger is activated to provide direct heating.

[0107] When the water temperature in the heat storage tank is in the medium temperature range and the user's water supply temperature is >55℃, the single-stage heat pump will be activated to adjust the temperature.

[0108] When the water temperature in the heat storage tank is in the low-temperature range but still higher than the ambient temperature, the two-stage compression heat pump is activated to improve the energy efficiency ratio.

[0109] When the water temperature in the heat storage tank is near the phase change stage and heat extraction needs to continue, the subcooling water generation function of the low-temperature stage evaporator of the two-stage heat pump is activated, allowing the formation of an ice-water mixture with a volume ratio not exceeding 10% in the tank, maximizing the extraction of sensible heat and part of latent heat.

[0110] The specific decision-making process is as follows: a fuzzy rule engine is used, with input variables including tank temperature, load demand, COP prediction value and electricity cost, and outputs including valve group opening and closing commands and heat pump start / stop / frequency conversion signals; for example: at 14:30 the system predicts that it will enter the medium temperature range and the user still needs 45℃ water supply, at this time the module automatically issues commands: close the plate heat exchanger bypass valve, open the single-stage heat pump supply loop valve, start the single-stage heat pump and set the target outlet water temperature to 48℃, and at the same time adjust the opening of the high temperature throttling valve to 70%.

[0111] IV. Edge Execution and Feedback Module:

[0112] The system receives generated control commands, drives pre-configured actuators, and monitors the deviation between the execution result and the target requirement in real time. If the deviation is within the corresponding standard threshold, no response is taken; otherwise (i.e., exceeding the limit), a secondary adjustment mechanism is triggered. The pre-configured actuators include: a PLC controller, an electric regulating valve drive unit, a frequency converter interface, and safety interlocks. In this embodiment, all valve actions are equipped with stroke limits and torque protection to prevent equipment damage due to misoperation. The monitoring process involves collecting actual operating parameters every L seconds after execution and comparing them with the target requirement. If the deviation exceeds the corresponding standard threshold, a secondary adjustment mechanism is triggered. Additionally, variables or parameters from the multi-stage heat exchange path decision module can be added to the actual operating parameters used here.

[0113] Wherein, the value of L is greater than 0, and in this embodiment, the value is usually set to 30;

[0114] The system sets the following allowable deviation range as the criterion for whether to trigger the secondary adjustment mechanism:

[0115] Upper / middle / lower water temperature deviation: Exceeds the set target value by ±2℃;

[0116] Flow deviation of inlet and outlet water pipes: exceeds the set value by ±5%;

[0117] Tank pressure deviation: Exceeds design pressure ±10% or falls below minimum static pressure limit;

[0118] User-side heat load response delay: The difference between actual heat supply and demand continues to exceed 10% for up to 60 seconds;

[0119] Compressor frequency response error: The difference between the set frequency and the measured frequency is > ±3Hz;

[0120] The temperature difference between the inlet and outlet of the evaporator / condenser deviates from the design value by ±1.5℃;

[0121] The aforementioned target values ​​and set values ​​can all be understood as target requirements, and will not be elaborated on here;

[0122] If any of the above parameters exceeds the corresponding standard threshold twice in a row, it is determined that the current control state deviates from the expectation, and the priority of the secondary adjustment mechanism is increased.

[0123] It should be noted that compatible actuators are used between the waste heat source and the heat user, including essential equipment for the heat pump process such as high-temperature throttling valves and low-temperature throttling valves. Other valves are used for switching between different operating conditions; please refer to the appendix for details. Figure 1 It can be seen that, under specific usage stages, it can be divided into heating mode and heat storage mode, which can be switched by valves as follows: Figure 2 As shown, the dashed lines indicate switching out of the current operating mode and having no effect. In heating mode, the heat pump system with plate heat exchanger extracts heat from the heat storage tank and supplies heat to users. In heat storage mode, the heat pump system extracts heat from the waste heat source to produce hot water, which is then sent to the heat storage tank. During heating operation, the water temperature is initially high, allowing for direct heat exchange via plate heat exchanger. As the temperature decreases, plate heat exchanger becomes insufficient for heating, necessitating single-stage heat pump heat exchange. As the temperature further decreases, the system switches to two-stage heat pump heat exchange. The low-temperature evaporator of the two-stage heat pump is capable of producing subcooled water, up to 10% ice-water mixture in the tank. The specific adjustment methods and underlying logic are the core of this solution. In heat storage mode, if the waste heat temperature is high or the tank temperature is low, direct heat exchange via plate heat exchanger can be used. If the waste heat temperature is low but the tank temperature is higher than the waste heat temperature, single-stage heat pump heat exchange can be used. If the waste heat temperature is very low or the tank temperature is very high, two-stage heat pump heat exchange is required. For details, please refer to [reference needed]. Figure 3 This is immediately apparent.

[0124] Effect Description: By adopting the above-described scheme, based on a multi-stage heat exchange path dynamic decision-making mechanism, it can automatically switch between direct plate heat exchanger supply, single-stage heat pump, two-stage heat pump, and even subcooled water generation modes according to the heat storage status, maximizing the matching of heat demand under different temperature zones and significantly improving the heat pump's coefficient of performance (COP). On the other hand, this decision-making process introduces a fuzzy rule engine, which integrates tank temperature, load demand, electricity price signals, and COP prediction values ​​for multi-objective optimization. Under the premise of ensuring stable heating supply, it prioritizes heat storage during off-peak electricity price periods, achieving optimal control of operating costs. This solves the problems of difficult heat extraction, frequent start-stop of heat pumps, and low energy utilization in traditional systems. The operation of the above scheme depends on the output results of the aforementioned temperature difference evolution model, providing instructions for precise adjustment at the execution layer and forming the core link of the system control chain.

[0125] The secondary adjustment mechanism is triggered by a three-level response strategy based on the type and magnitude of the deviation:

[0126] Level 1 regulation: Applicable to situations where the deviation is small and the system structure has not failed. It implements proportional regulation based on the deviation amount. If the water supply temperature deviation exceeds the limit, the compressor frequency is adjusted based on: Δf=K. f ×ΔT; where Δf is the compressor frequency adjustment; ΔT represents the water supply temperature deviation; K f The frequency adjustment proportional coefficient, in Hz / ℃, ranges from 2.0 to 3.0 Hz / ℃, with a preferred value of 2.5 Hz / ℃ in this embodiment. The above formula is based on an approximately linear relationship between the heat pump's heating capacity and the compressor's frequency. Experimental calibration shows that under rated operating conditions, for every 1 Hz increase in operating frequency, the heat pump's heating capacity increases by approximately 6-9 kW, depending on the model and operating conditions. When the actual water supply temperature is detected to be lower than the set value, the system proportionally increases the frequency to enhance heating capacity. The final result is taken upwards to avoid frequent fine-tuning. The new frequency obtained after adjustment can be the sum of the original and the adjusted value, but it must not exceed the heat pump's maximum allowable frequency. If the flow deviation in the inlet and outlet water pipes exceeds the limit, the circulating pump's frequency conversion adjustment action is executed, based on:

[0127] Δf p =K p ×|(G set -G meas ) / G set |×100%; where Δf p G represents the percentage adjustment of the output frequency of the circulating pump motor inverter. set For the target loop flow, G meas K represents the measured flow rate of the inlet and outlet water pipes. p The flow rate adjustment gain coefficient is dimensionless and ranges from 0.7 to 1.0; in this embodiment, it is 0.8. The above formula adopts a proportional control strategy to ensure that the flow rate deviation converges quickly. Since the pump power is proportional to the cube of the speed, large adjustments are avoided to prevent a surge in energy consumption. After actual adjustment, continuous monitoring is required until the deviation does not exceed ±5% of the set value. If the actual heat exchange temperature difference of the evaporator is less than 75% and the condenser output is insufficient, i.e., the temperature difference between the inlet and outlet of the evaporator / condenser deviates from the design value by ±1.5℃, then the heat exchange surface is determined to be contaminated or scaled. The disturbance cleaning program is started: the dedicated cleaning solenoid valve is opened for 10 seconds to introduce a high-speed water pulse to flush the plate heat exchange channel and improve the heat exchange efficiency. This action is performed at most once every 24 hours to prevent excessive disturbance from affecting system stability.

[0128] Secondary regulation: When the water temperature continues to drop to the threshold of the next temperature zone, for example, when moving from the medium temperature zone to the low temperature zone, or when the heat pump COP is lower than the set value and lasts for 5 minutes, the trend analysis and identification module is called again to identify the status and request the multi-stage heat exchange path decision module to re-determine the heat exchange path; for example, if a single-stage heat pump is used for heating, the energy efficiency drops sharply due to the tank temperature dropping to 38°C, and the system automatically switches to a two-stage heat pump mode; if a sudden increase of 30% in user load is detected, the system starts the backup heat pump in advance and increases the flow rate set value according to the predetermined step value.

[0129] Three-level regulation: If the lower water temperature is in the near-phase change stage and continues to drop, or the flow rate in the inlet and outlet water pipes is zero for more than 10 seconds, then immediately execute the protection action: shut down the heat pump unit, open the bypass valve, and start the antifreeze circulation pump to prevent ice blockage or dry burning.

[0130] Effect Description: By setting up a three-level response secondary adjustment mechanism, graded intervention can be implemented when operational deviations occur: the first-level adjustment performs proportional compensation based on the deviation amount to maintain stable system operation; the second-level adjustment re-invokes the trend identification and path decision module when operating conditions change abruptly, realizing dynamic reconstruction of the control strategy; the third-level adjustment executes emergency protection actions in extreme cases to prevent safety accidents such as ice blockage and dry burning; at the same time, the operation of this mechanism realizes quantitative adjustment, avoiding the overshoot and oscillation problems that are prone to occur in traditional PID control, improving adjustment accuracy and system robustness, and solving the problems of coarse adjustment, lack of hierarchical response and insufficient safety of traditional control systems. The operation of this scheme directly depends on the state identification results and decision instructions provided in the two effect descriptions given above.

[0131] In addition, the module includes integrated emergency shutdown logic: when the tank pressure exceeds the limit, water flow stagnates, or there is a risk of ice blockage, the heat pump power is immediately cut off and the emergency bypass is activated; for example, after the valve switching is completed at 14:35, the system monitors that the actual supply water temperature is 47.8℃, the return water temperature is 40.2℃, and the flow rate is stable at 118m³ / h. 3 If the temperature is within the set range, the execution is considered successful; if the temperature is only 43℃, the heat pump compressor frequency will be automatically increased by 5% until the set value is reached.

[0132] In summary, this embodiment provides a holistic example of running the system:

[0133] Taking a typical winter heating cycle as an example: For instance, at 06:00, the system detects that the heat storage tank is full and the water temperature is 95℃, and users begin to use heat; the data sensing and acquisition module collects initial data, the trend analysis and identification module identifies it as a high-temperature segment, the multi-stage heat exchange path decision module decides to activate direct supply from the plate heat exchanger, and the edge execution and feedback module executes the opening of the plate heat exchanger valve group; by 12:00, the tank temperature drops to 58℃, the trend analysis and identification module warns that it is about to enter the medium-temperature segment, the multi-stage heat exchange path decision module prepares the single-stage heat pump start-up sequence in advance, and automatically switches to [heat exchanger] at 12:15. Heat pump heating; during off-peak electricity hours at night, around 23:00, the waste heat source temperature is only 30℃, while the water temperature inside the tank has risen to 70℃, making conventional heat exchange impossible. The multi-stage heat exchange path decision module decides to activate the two-stage heat pump reverse heat storage, injecting the low-temperature waste heat into the water tank. At 07:00 the next day, the tank temperature drops to 8℃ again, and the trend analysis and identification module activates the phase change early warning mechanism. The multi-stage heat exchange path decision module then activates the two-stage heat pump deep heat extraction mode, allowing the generation of an 8% ice-water mixture, thus extending the effective heat extraction lower limit from the traditional 5℃ to close to 0℃.

[0134] Therefore, by adopting the above technical solution, the problems of large storage tank volume, large footprint, and high investment caused by the limited temperature difference in traditional water thermal storage systems are solved. At the same time, it provides an effective means to deal with the difficulty of heat extraction in the traditional low-temperature section, avoiding heat waste caused by low-temperature water and resulting in low overall system energy efficiency. The system uses edge computing to identify the temperature difference evolution trend in real time and dynamically switch multi-stage heat exchange paths, realizing ultra-large temperature difference heat storage and release from 95℃ to near 0℃, reducing the water tank volume by more than 45% for the same heat storage capacity. On the other hand, by using edge intelligence to accurately control the ice-water mixture ratio in the near-phase change zone, it not only avoids the risk of pipe freezing and blockage, but also significantly improves the heat storage density per unit water volume, achieving a non-steady-state high-density heat storage effect that conventional water thermal storage systems cannot achieve.

[0135] Example 2:

[0136] Based on Example 1, this example also provides a thermal storage management method based on edge computing and temperature difference information, including: acquiring the operating parameters of the target thermal storage system and forming a basic dataset after preprocessing;

[0137] Based on the basic dataset, a temperature gradient evolution model is constructed inside the target thermal storage system to identify the current thermal storage state and predict the evolution inference results.

[0138] Based on the output evolution reasoning results, combined with the external waste heat source temperature, user-side water supply requirements and electricity price time-of-use signals, the optimal heat exchange path and heat pump operating stage are dynamically selected to generate control commands.

[0139] It receives generated control commands, drives the pre-configured actuators to perform actions, and monitors the deviation between the execution results and the target requirements in real time. When the deviation is outside the corresponding standard threshold, a secondary adjustment mechanism is triggered.

[0140] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0141] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0142] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A thermal storage management system based on edge computing and temperature difference information, characterized in that, include: The data sensing and acquisition module is used to acquire operating parameters, including the water temperature of the upper, middle and lower layers of the thermal storage tank, which are preprocessed to form a basic dataset. The trend analysis and identification module is used to construct a temperature gradient evolution model inside the heat storage tank based on the basic dataset, identify the temperature range to which the current heat storage state belongs, and output the evolution inference results including the expected time to enter the next temperature range. The multi-stage heat exchange path decision module is used to dynamically select the optimal heat exchange path among plate heat exchanger direct supply, single-stage heat pump or double-stage heat pump based on the evolution reasoning results, combined with user-side water supply requirements and electricity price time-of-use signals, so as to generate path switching instructions. The edge execution and feedback module is used to receive the path switching command, drive the actuators including the electric regulating valve and the frequency converter pump to move, and monitor the deviation between the execution result and the target requirement in real time. When the deviation exceeds the limit, a secondary adjustment mechanism is triggered. A temperature zone synchronization control signal generator is used to generate a periodic control signal whose period is associated with the expected time to enter the next temperature zone; The edge execution and feedback module is further configured to: during the transition from the current temperature zone to the next temperature zone, perform cyclic driving and state sampling of the actuator in the currently activated heat exchange path according to the control signal, and change the average output power of the actuator by adjusting the pulse width or frequency of the control signal, so as to achieve precise stabilization of water supply temperature or flow rate.

2. The thermal storage management system based on edge computing and temperature difference information according to claim 1, characterized in that, The control signal is a pulse density modulation signal, whose pulse density is proportional to the required actuator action intensity, and is used to achieve gradual adjustment of the opening of the electric regulating valve or smooth control of the speed of the variable frequency pump.

3. The thermal storage management system based on edge computing and temperature difference information according to claim 1, characterized in that, Each cycle of the control signal is divided into an energy injection period, a system response period, and a state sampling period; During the energy injection period, full drive power is output to the actuator to rapidly change its state; During the system response period, the driving power is stopped or reduced, and the thermodynamic system is allowed to reach a new equilibrium under its own inertia. During the state sampling period, key parameters are collected to evaluate the control effect and determine the signal parameters for the next cycle. The duration of the system response period is dynamically set based on the thermal inertia of the water in the thermal storage tank and the type of heat exchange path currently in operation, to ensure that the system has stabilized before the state sampling. The system is configured to: calculate the instantaneous heat transfer efficiency index during the state sampling period, and dynamically adjust the duration of the energy injection period in the next cycle based on the deviation of the index from the expected value, so as to achieve closed-loop optimization of the heat transfer process.

4. The thermal storage management system based on edge computing and temperature difference information according to claim 1, characterized in that, The generation period of the temperature zone synchronization control signal is determined by the expected time to enter the next temperature zone and the thickness of the current temperature zone; the shorter the expected transition time or the smaller the temperature zone thickness, the higher the frequency of the signal and the finer the control granularity.

5. The thermal storage management system based on edge computing and temperature difference information according to claim 1, characterized in that, The drive circuit of the actuator integrates a current feedback unit; the edge execution and feedback module indirectly determines whether the valve is stuck or the pump is running dry by monitoring the change of the drive current, and uses this information as a feedback signal to adjust the control signal.

6. The thermal storage management system based on edge computing and temperature difference information according to claim 1, characterized in that, The operating parameters of the temperature zone synchronous control signal generator are modulated by the phase change early warning signal output by the trend analysis and identification module. When a phase change early warning occurs, the signal frequency is automatically increased and the intensity of a single energy injection is reduced to achieve fine control against freezing and blockage in the near phase change zone.

7. The thermal storage management system based on edge computing and temperature difference information according to claim 1, characterized in that, The multi-stage heat exchange path decision module generates a corresponding initial control strategy while outputting the path switching command; the temperature zone synchronous control signal generator initializes the frequency and pulse width of the signal based on this strategy.

8. The thermal storage management system based on edge computing and temperature difference information according to claim 1, characterized in that, The secondary adjustment mechanism works in conjunction with the temperature zone synchronous control signal: First-level regulation involves real-time compensation through fine-tuning the control signal parameters of the current cycle; Secondary regulation: When primary regulation fails and the water temperature is about to cross the temperature zone boundary, the trend analysis and identification module is triggered to make a new prediction and request the multi-stage heat exchange path decision module to make a new decision. The three-stage regulation system, when it detects that the lower water temperature has entered the near-phase change zone and continues to decline, forcibly overrides the current regulation signal and executes an emergency protection action to shut down the valve.

9. A thermal storage management method based on edge computing and temperature difference information, characterized in that, Includes the following steps: The operating parameters, including the water temperatures of the upper, middle, and lower layers of the thermal storage tank, are obtained and preprocessed to form a basic dataset. Based on the aforementioned basic dataset, a temperature gradient evolution model inside the thermal storage tank is constructed to identify the temperature range to which the current thermal storage state belongs and to predict the evolutionary reasoning results for the time to enter the next temperature range. Based on the evolutionary reasoning results, combined with the user-side water supply requirements and electricity price time-of-use signals, the optimal heat exchange path among plate heat exchanger direct supply, single-stage heat pump or two-stage heat pump is dynamically selected to generate a path switching command. Upon receiving the path switching instruction, the pre-configured actuator is driven to perform the action, and the deviation between the execution result and the target requirement is monitored in real time. When the deviation is outside the corresponding standard threshold, a secondary adjustment mechanism is triggered. This also includes: A periodic control signal is generated, the period of which is synchronized with the temperature transition time predicted based on temperature difference information; During the temperature transition, the actuators in the currently activated heat exchange path are cyclically driven and their status is sampled based on this control signal. By adjusting the pulse width or frequency of the control signal, the average output power of the actuator is changed, thereby achieving precise stabilization of the water supply temperature or flow rate.

10. The thermal storage management method based on edge computing and temperature difference information according to claim 9, characterized in that, The pulse density modulation signal is used as the control signal, and the average output power of the actuator is linearly controlled by changing the number of pulses per unit time. Each cycle of the control signal is divided into an energy injection period, a system response period, and a state sampling period to match the driving force, inertial response, and evaluation program of the thermodynamic system, thereby achieving stable control. During the state sampling period of each control signal cycle, an instantaneous energy efficiency coefficient is calculated based on the collected temperature and flow data and compared with the target energy efficiency coefficient. Based on this, the intensity of the energy injection period in the next cycle is adjusted to achieve dynamic energy efficiency optimization. The phase change warning level is mapped to a dual constraint on the frequency and pulse width of the control signal. Under high risk, a high-frequency, low-pulse-width control mode is adopted to balance heat extraction efficiency and antifreeze safety.

Images