Waste heat efficient recovery and intelligent distribution method and system

Abstract

The invention provides a waste heat efficient recovery and intelligent distribution method and system, and belongs to the technical field of waste heat recycling. The method comprises the steps that operation parameters of a compressor, heat supply medium temperature data and the heat supply requirement of a target area are obtained; inputting the operation parameters and the heat supply medium temperature data into a pre-trained waste heat prediction model to generate a waste heat prediction value of recoverable waste heat; calculating a heat supply demand value of the target area according to the heat supply demand of the target area and a thermodynamic transfer model; and determining a corresponding distribution strategy according to the waste heat prediction value and the heat supply demand value. According to the waste heat efficient recovery and intelligent distribution method and system, accurate prediction and reasonable distribution of waste heat are achieved by collecting and analyzing the operation parameters of the compressor, the temperature data of the heat supply medium and the heat supply demand of the target area, the waste heat recovery and utilization efficiency is improved, and the heat supply demand of the target area is met.

Description

Technical Field

[0001] The present application relates to the technical field of waste heat recovery and utilization, and in particular to a method and system for efficient waste heat recovery and intelligent distribution. Background Art

[0002] During industrial production, equipment such as compressors generates significant amounts of waste heat. Traditional waste heat recovery methods suffer from low efficiency and irrational distribution. Existing technologies often fail to accurately predict the amount of recoverable waste heat and intelligently allocate it based on actual heating needs. This results in significant waste of waste heat resources and fails to meet the needs for efficient and stable heating in target areas.

[0003] Therefore, there is an urgent need for a waste heat efficient recovery and intelligent distribution method and system method that can achieve accurate recovery and reasonable distribution of waste heat, so as to improve energy utilization efficiency and reduce energy consumption and production costs. Summary of the Invention

[0004] In order to solve the above technical problems, the present application provides a method and system for efficient recovery and intelligent distribution of waste heat. By collecting and analyzing the compressor operating parameters, heating medium temperature data and heating demand of the target area, accurate prediction and reasonable distribution of waste heat can be achieved, thereby improving the efficiency of waste heat recovery and utilization and meeting the heating demand of the target area.

[0005] A first aspect of an embodiment of the present application provides a method for efficient waste heat recovery and intelligent distribution, comprising: Obtain the compressor's operating parameters, heating medium temperature data, and target area heating demand; Inputting the operating parameters and the heating medium temperature data into a pre-trained waste heat prediction model to generate a waste heat prediction value of recoverable waste heat; Calculating the heating demand value of the target area according to the heating demand of the target area and the thermodynamic transfer model; A corresponding allocation strategy is determined according to the waste heat prediction value and the heating demand value.

[0006] A second aspect of the embodiments of the present application provides a waste heat efficient recovery and intelligent distribution system, comprising: Data acquisition module, used to obtain the operating parameters of the compressor, the temperature data of the heating medium and the heating demand of the target area; a data prediction module, configured to input the operating parameters and the heating medium temperature data into a pre-trained waste heat prediction model to generate a waste heat prediction value of recoverable waste heat; A data calculation module is used to calculate the heating demand value of the target area according to the heating demand of the target area and a thermodynamic transfer model; A strategy determination module is used to determine a corresponding allocation strategy based on the waste heat prediction value and the heating demand value.

[0007] According to a third aspect of an embodiment of the present application, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein when the processor executes the computer program, the steps of the above-mentioned method for efficient recovery and intelligent distribution of waste heat are implemented.

[0008] In a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned method for efficient recovery and intelligent distribution of waste heat are implemented.

[0009] In summary, the beneficial effects of a method and system for efficient recovery and intelligent distribution of waste heat provided by the embodiment of the present application are as follows: the waste heat prediction model obtained by the present application based on historical data and neural network modeling can accurately predict future recoverable waste heat, provide a reliable basis for reasonable distribution, and improve the efficiency of waste heat recovery. Secondly, combined with the multi-dimensional heating demand and thermodynamic transfer model of the target area, the heating demand value is accurately calculated by comprehensively considering factors such as the environment, buildings and user settings to ensure the heating effect. In determining the distribution strategy, the distribution strategy of waste heat is determined based on the matching degree between the waste heat prediction value and the heating demand value, which realizes the intelligent distribution and improves the efficiency of waste heat recovery and utilization. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 A schematic diagram of a flow chart of a method for efficient waste heat recovery and intelligent distribution provided in one embodiment of the present application; Figure 2 This is a structural block diagram of a waste heat efficient recovery and intelligent distribution system provided in one embodiment of the present application; Figure 3 A schematic diagram of the structure of a multi-stage heat exchanger provided in one embodiment of the present application.

[0011] Figure 4 A schematic block diagram of an electronic device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0012] In the following description, specific details such as specific system structures and techniques are provided for purposes of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present application with unnecessary detail.

[0013] To make the purpose, technical solutions and advantages of this application clearer, Figure 1-4 The following description will be given through specific examples.

[0014] Please refer to Figure 1 , Figure 1 A schematic flow chart of a method for efficient waste heat recovery and intelligent distribution provided in one embodiment of the present application, the method comprising: S101: Obtaining compressor operating parameters, heating medium temperature data, and target area heating demand.

[0015] In this embodiment, the operating parameters of the compressor are obtained by selecting a pressure sensor, a speed sensor and a power sensor, which are respectively installed at key positions such as the air inlet, exhaust port, rotating shaft and power supply line of the compressor. The sensor acquisition frequency is set to 1 time per second, and the collected data is transmitted to the data processing center via industrial Ethernet.

[0016] Acquisition of heating medium temperature data: armored thermocouple temperature sensors are installed at the inlet and outlet of the heating medium pipeline and at nodes at every first interval (for example, 100 meters), and the temperature data is sent to the data processing center through a wireless transmission module.

[0017] Obtaining heating demand in the target area: Install intelligent data collection terminals in each building in the target area to collect information such as the building area, usage function, occupant density, indoor set temperature, etc. of each building in the target area; in addition, establish a data interface with the meteorological department to obtain meteorological data (such as outdoor temperature, humidity, wind speed, etc.) in real time.

[0018] S102: Inputting the operating parameters and the heating medium temperature data into a pre-trained waste heat prediction model to generate a waste heat prediction value of recoverable waste heat.

[0019] In this embodiment, the method for constructing the pre-trained waste heat prediction model includes: Collect historical compressor operating parameters, heating medium temperature data and actual recoverable waste heat data, and divide them into training set, validation set and test set according to the proportion; if a large amount of data is not collected, the generative adversarial network model can be used to expand the above collected historical data.

[0020] A waste heat prediction model was constructed based on a deep learning algorithm (long short-term memory network algorithm). Three LSTM layers were configured, with 64, 32, and 16 neurons in each layer, respectively. The output layer had one neuron, which outputted the waste heat prediction value. The mean squared error (MSE) loss function was selected. Training was performed for 100 epochs on the training set. Model performance was monitored and hyperparameters were adjusted on the validation set, and the model's accuracy and generalization were finally verified on the test set. During the training phase, the waste heat prediction model of this embodiment utilized historical operating parameters, heating medium temperature data, and corresponding actual recoverable waste heat data. The model parameters were adjusted using a backpropagation algorithm to learn the complex mapping relationship between the three, thereby generating a waste heat prediction value for recoverable waste heat.

[0021] S103: Calculate the heating demand value of the target area based on the heating demand of the target area and the thermodynamic transfer model.

[0022] In this example, a thermodynamic transfer model is established based on the target area's heating network topology, pipe material, diameter, length, and other parameters, combined with fundamental thermodynamic laws (such as Fourier's law and the law of conservation of energy). This model accurately describes the heat transfer process within the heating network, including conduction, convection, and radiation.

[0023] In this embodiment, the heating demand of the target area (heating demand includes: actual user demand and predicted demand) is input into the thermodynamic transfer model, and factors such as heat loss in the heating pipe network and heat transfer delay are used to calculate the heating demand value of the target area; wherein the heating demand value includes the specific heating load demand in the area.

[0024] S104: Calculate the heating demand value of the target area based on the heating demand of the target area and the thermodynamic transfer model; and determine a corresponding allocation strategy based on the waste heat prediction value and the heating demand value.

[0025] In this embodiment, a corresponding allocation strategy is determined based on the predicted residual heat value and the heating demand value. When the predicted residual heat value is greater than the heating demand value, the heating demand of the target area is prioritized, and the excess residual heat is stored in the thermal storage device. If the predicted residual heat value is less than the heating demand value, the residual heat is allocated according to the pre-set heating priority of each area (determined by factors such as regional importance and user type), and auxiliary heat sources are activated to supplement the heat gap. When the predicted residual heat value is close to the heating demand value, a dynamic balance allocation strategy is adopted to adjust the residual heat allocation ratio in real time to ensure stable operation of the heating system. The dynamic balance allocation strategy uses a PID controller to adjust the opening adjustment of the residual heat distribution valve and the speed adjustment of the residual heat distribution pump in each area in real time to achieve dynamic and balanced distribution of residual heat.

[0026] From the above, it can be concluded that this application achieves accurate prediction of waste heat through multi-source data collection and intelligent algorithms, providing a reliable basis for the rational allocation of waste heat and improving the accuracy and reliability of waste heat recovery. Based on the actual heating demand of the target area and the thermodynamic transfer model, the heating demand value is calculated, and a variety of allocation strategies are formulated in combination with the waste heat prediction value. This can meet the heating demand of the target area and significantly improve the efficiency of waste heat utilization. This application also helps to reduce energy consumption and production costs in industrial production processes.

[0027] In one embodiment of the present application, a training method for a pre-trained residual heat prediction model includes: Collect the compressor operating parameter sequence, heating medium temperature time series data and the corresponding actual waste heat recovery amount within the historical time period; The operating parameter sequence and heating medium temperature time series data are used as input features, and the actual waste heat recovery amount is used as the label. The time series modeling training is performed through the neural network model to obtain the waste heat prediction model.

[0028] In this embodiment, a sequence of operating parameters of the compressor within a historical time period is collected, including but not limited to the compressor's speed, power, pressure, flow, etc., to form multidimensional time series data; the time series data of the heating medium temperature within the corresponding time period is simultaneously collected, and the temperature change curve is obtained from multiple monitoring points; the actual waste heat recovery amount corresponding to each time point is recorded as a supervisory label for model training.

[0029] The collected raw data is time-aligned to ensure that the operating parameter sequence, temperature time series data and the actual waste heat recovery amount correspond one-to-one in the time dimension; the operating parameter sequence and heating medium temperature time series data are sampled using a sliding window to generate fixed-length time series samples.

[0030] Calculate statistical features within each time window, such as mean, variance, maximum, minimum, etc., to enhance the model's ability to express the characteristics of time series data; extract the periodic and trend characteristics of time series data.

[0031] The time series data is modeled based on a deep learning algorithm (long short-term memory network algorithm) to obtain a residual heat prediction model. Furthermore, this embodiment can also introduce an attention mechanism into the long short-term memory network algorithm, enabling the model to automatically focus on key parts of the time series data, thereby improving prediction accuracy.

[0032] In one embodiment of the present application, the heating demand value of the target area is calculated based on the heating demand of the target area and the thermodynamic transfer model, including: The heating power of the target area is calculated based on the heating demand of the target area and the thermodynamic transfer model; The heating demand includes: the ambient temperature of the target area, the building thermal resistance coefficient of the target area, and the user-set temperature of the target area; The temperature fluctuation compensation amount is calculated based on the historical temperature change rate of the target area and the predicted temperature difference. The predicted temperature difference is the difference between the user-set temperature of the target area and the predicted ambient temperature of the target area.

[0033] In this embodiment, the heating demand includes multiple key parameters such as the ambient temperature of the target area, the building thermal resistance coefficient of the target area, and the user-set temperature of the target area. The ambient temperature of the target area represents the outdoor temperature under external natural conditions and is basic data that changes in real time; the building thermal resistance coefficient of the target area is the thermal insulation performance of the building itself. The thermal resistance coefficients of different building structures and materials vary significantly, which directly affects the heat transfer loss; the user-set temperature of the target area is the user's expectation of a comfortable indoor temperature and is the core goal orientation of the operation of the heating system. This embodiment calculates heating power using thermodynamic principles such as Fourier's law of heat conduction and Newton's cooling equation, in conjunction with a thermodynamic transfer model. In one embodiment, based on the total area of ​​buildings within the target area, the building thermal resistance coefficient, and the difference between the ambient temperature and the user-set temperature, the formula Q = K × A × ΔT is used, where Q is the heat flux, K is the total heat transfer coefficient (related to the building thermal resistance coefficient), A is the heat transfer area, and ΔT is the temperature difference. This allows for a preliminary estimate of the base heating power required to maintain the target area at the user-set temperature. In this embodiment, to more accurately meet heating demand, a temperature fluctuation compensation is introduced. This compensation is calculated based on a weighted combination of the target area's historical temperature change rate and the predicted temperature difference. The predicted temperature difference is the difference between the target area's user-set temperature and the target area's predicted ambient temperature. The predicted ambient temperature can be analyzed and predicted using weather forecast data from meteorological authorities, combined with regional climate characteristics and temporal patterns. In this embodiment, the historical temperature change rate is calculated by collecting temperature change data from the target area over a period of time (e.g., the past week or month) and calculating the temperature change amplitude per unit time. For example, using time series analysis methods, the historical temperature data is smoothed and the ratio of the temperature difference between adjacent time points to the time interval is calculated to obtain the average temperature change rate. In this embodiment, when the temperature fluctuation compensation amount is calculated using the first formula, the first formula is: ; Among them, ΔQ is the temperature fluctuation compensation amount, T is the historical temperature change rate, ΔT is the predicted temperature difference, α is the first weighting coefficient, β is the second weighting coefficient, and α + β = 1; the superscript T represents transpose. In actual application scenarios, if the temperature change in the target area is relatively stable, the second weighting coefficient β of the predicted temperature difference can be appropriately increased; if the temperature in the area fluctuates frequently, the first weighting coefficient α of the historical temperature change rate is increased, so as to achieve dynamic compensation for temperature fluctuations, enabling the heating system to respond to environmental changes more timely and accurately and ensuring the heating effect in the target area.

[0034] In this embodiment, by adding the basic heating power calculated based on the thermodynamic transfer model and the power corresponding to the temperature fluctuation compensation amount, the final heating demand value (Q + ΔQ) of the target area can be obtained. This embodiment realizes precise regulation of heating and efficient and energy-saving heating operation.

[0035] In an embodiment of the present application, a corresponding allocation strategy is determined according to the waste heat prediction value and the heating demand value, including: Calculate the matching degree between the waste heat prediction value and the heating demand value; When the matching degree is greater than or equal to the first matching degree threshold, select the corresponding allocation strategy in the allocation strategy library according to the matching degree; When the matching degree is less than the first matching degree threshold, based on the reinforcement learning algorithm, process the operating parameters of the compressor, the temperature data of the heating medium, and the heating demand of the target area to determine the allocation strategy.

[0036] In this embodiment, the matching degree M between the waste heat prediction value and the heating demand value is calculated through the relative error. The value range of M is [0, 1], and the larger the value, the better the matching.

[0037] When the matching degree is greater than or equal to the first matching degree threshold T1, enter the rule-based strategy selection mode: if M ≥ T1, then S = SelectStrategy(M, StrategyDB), where the strategy library (StrategyDB) is stored in a multi-dimensional index structure and includes the following dimensions: matching degree interval, time period characteristics, waste heat type, demand priority (residential / commercial / industrial users), compressor start-stop control, and opening degree of the flow regulating valve, etc.

[0038] When the matching degree is less than the first matching degree threshold, the system switches to the reinforcement learning mode: If M < T1, then S = RLModel(Ot, At, Rt), where the state space Ot includes: operating parameters of the compressor (frequency, power, etc.), temperature data of the heating medium (supply water / return water temperature), heating demand of the target area; The action space At includes: compressor start-stop control, opening degree of the flow regulating valve, charge / discharge control of the energy storage device.

[0039] This application, through this two-tier decision-making mechanism, ensures allocation efficiency while having the adaptive ability to cope with complex working conditions, and realizes the intelligent optimization allocation of waste heat resources.

[0040] In this embodiment, a second matching degree threshold T2 is further set. When the matching degree is continuously less than the second matching degree threshold, an emergency response mechanism is triggered, including starting a backup heat source, adjusting energy storage equipment, etc.

[0041] In one embodiment of the present application, the operating parameters of the compressor, the temperature data of the heating medium, and the heating demand of the target area are processed based on a reinforcement learning algorithm to determine an allocation strategy, including: Determine the initial action space vector of the reinforcement learning algorithm corresponding to the allocation strategy; Determine multiple standard allocation strategies and determine the action space vector of the reinforcement learning algorithm corresponding to each standard allocation strategy; Determine the state space vector of the reinforcement learning algorithm based on the compressor operating parameters, heating medium temperature data, and target area heating demand; determining a discount factor based on a transmission delay from the compressor to the target area; Determine the value function of the reinforcement learning algorithm based on the action space vector, state space vector, and discount factor; The allocation strategy is determined based on the initial action space vector and the feedback value obtained from the value function.

[0042] In this embodiment, the initial action space vector of the reinforcement learning algorithm corresponding to the allocation strategy is determined , which has the following dimensions:

[0043] in, is the compressor frequency adjustment coefficient; is the opening of the heating medium flow regulating valve, Charge and discharge power for energy storage equipment; is the district heating priority weight vector.

[0044] In this embodiment, multiple standard allocation strategies are determined. , and construct its mapping relationship with the action space vector:

[0045] Among them, the mapping function It is obtained through calibration of expert knowledge and historical data.

[0046] In this embodiment, the state space vector of the reinforcement learning algorithm is determined based on the operating parameters P of the compressor, the temperature data of the heating medium T0 and the heating demand D of the target area.

[0047] Among them, P = (compressor frequency, power, current); T0 = (supply water temperature, return water temperature, ambient temperature, temperature prediction gradient); D = (demand in each region, demand change trend); H = (energy storage level, transmission status, waste heat prediction).

[0048] In this embodiment, the state vector uses a sliding window mechanism to construct time series features.

[0049] In this embodiment, the discount factor is determined based on the transmission delay τ from the compressor to the target area. :

[0050] in, is the basic discount factor, is the delay attenuation coefficient, determined according to the transmission pipeline; τ is the transmission delay, and ; Where L is the length of the pipeline, v is the flow velocity of the medium, Heat transfer delay.

[0051] Based on the action space vector , state space vector and discount factor Determining the value function of a reinforcement learning algorithm :

[0052] in, is the value function, is the mathematical expectation operator; For the time step The reward value obtained when For the moment The transmission delay from the compressor to the target area, is the discount factor, which is about the transmission delay The function is used to discount the rewards at different time steps in the future, making the recent rewards more important than the distant rewards.

[0053] In another embodiment, the method further includes: determining a value function of a reinforcement learning algorithm based on the action space vector, the state space vector, and the discount factor.

[0054] in, is the value of the reward function, is the maximum Q value in the future state, is the discount factor, is the new action space vector, is the new state space vector.

[0055] Among them, the reward function can be designed as a comprehensive indicator, including: heating satisfaction (the deviation between the target area temperature and the set temperature), energy consumption efficiency (the ratio of compressor power to heating effect), system stability (parameter fluctuation amplitude), etc.

[0056] This application, through a transmission delay-aware reinforcement learning framework, can dynamically adapt to the characteristics of the heating network, optimize energy utilization efficiency while meeting user needs, and realize the intelligent allocation of waste heat resources.

[0057] In one embodiment of the present application, the operating parameters of the compressor, the temperature data of the heating medium, and the heating demand of the target area are processed based on a reinforcement learning algorithm to determine the allocation strategy, which also includes: Adjust the discount factor based on the transmission delay from the compressor to the target area; Determine the value function of the reinforcement learning algorithm based on the action space vector, state space vector, and discount factor; The discount factor is adjusted based on the transmission delay from the compressor to the target area, including: In response to the transmission delay being less than a first threshold, reducing the discount factor by a first step; In response to the transmission delay being greater than a second threshold, the discount factor is increased by a second step size, and the first deviation threshold is less than the second deviation threshold.

[0058] In this embodiment, when it is detected that the transmission delay is less than the preset first threshold, it indicates that the heating response is rapid and the system has good real-time performance; at this time, the discount factor is reduced by the first step; the lower discount factor enables reinforcement learning to focus more on immediate rewards and optimize short-term heating efficiency.

[0059] When the transmission delay is greater than the preset second threshold, it means that the heating response is lagging and it is necessary to strengthen the consideration of future states. At this time, the discount factor is increased by the second step. A higher discount factor makes the algorithm pay more attention to long-term cumulative rewards and avoid frequent adjustments due to short-term fluctuations.

[0060] When the transmission delay is between the first threshold and the second threshold, the discount factor is fine-tuned using linear interpolation or a sigmoid function to achieve a smooth transition.

[0061] This application uses a reinforcement learning method that dynamically adjusts the discount factor based on transmission delay, which can better adapt to the time-varying characteristics of the heating system, balance short-term heating effects and long-term system stability, and achieve more accurate and efficient heating resource allocation.

[0062] In this embodiment, the transmission delay from the compressor to the target area is These include: fluid transport delays, heat transfer delays, and delays caused by thermal storage devices.

[0063] The method for determining the discount factor further includes: determining the discount factor according to a second formula:

[0064] in, is the basic discount factor, is the delay attenuation coefficient, is the first adjustment parameter, is the second adjustment parameter, is the reference delay threshold.

[0065] The transmission delay-aware discount factor dynamic adjustment mechanism proposed in this application can effectively improve the application effect of reinforcement learning in regional heating systems, especially in long-distance transmission scenarios, thereby improving energy utilization efficiency.

[0066] In one embodiment of the present application, the method for efficient waste heat recovery and intelligent distribution further includes: when the waste heat prediction value is less than the heating demand value, selecting a multi-source collaborative heating mode in the distribution strategy library.

[0067] In this embodiment, the multi-source collaborative mode in the allocation strategy library is intelligently combined through the heat source priority matrix, as shown in Table 1.

[0068] Table 1 Heat source priority matrix

[0069] This application sets up an intelligent coordination mechanism to maintain efficient, economical and reliable heating services when there is insufficient waste heat.

[0070] Corresponding to a waste heat efficient recovery and intelligent distribution method in the above embodiment, Figure 2 This is a structural block diagram of a waste heat efficient recovery and intelligent distribution system provided by an embodiment of the present application. For ease of explanation, only the parts related to the embodiment of the present application are shown. Figure 2 The waste heat efficient recovery and intelligent distribution system 20 includes: a data acquisition module 21, a data prediction module 22, a data calculation module 23 and a strategy determination module 24.

[0071] The data acquisition module 21 is used to obtain the operating parameters of the compressor, the temperature data of the heating medium and the heating demand of the target area; The data prediction module 22 is used to input the operating parameters and the temperature data of the heating medium into the pre-trained waste heat prediction model to generate a waste heat prediction value of recoverable waste heat; A data calculation module 23 is used to calculate the heating demand value of the target area based on the heating demand of the target area and the thermodynamic transfer model; The strategy determination module 24 is used to determine a corresponding allocation strategy according to the waste heat prediction value and the heating demand value.

[0072] In one embodiment of the present application, the data prediction module 22 is specifically configured to: Collect the compressor operating parameter sequence, heating medium temperature time series data and the corresponding actual waste heat recovery amount within the historical time period; The operating parameter sequence and heating medium temperature time series data are used as input features, and the actual waste heat recovery amount is used as the label. The time series modeling training is performed through the neural network model to obtain the waste heat prediction model.

[0073] In one embodiment of the present application, the data calculation module 23 is specifically configured to: The heating power of the target area is calculated based on the heating demand of the target area and the thermodynamic transfer model; The heating demand includes: the ambient temperature of the target area, the building thermal resistance coefficient of the target area, and the user-set temperature of the target area; The temperature fluctuation compensation amount is calculated based on the historical temperature change rate of the target area and the predicted temperature difference. The predicted temperature difference is the difference between the user-set temperature of the target area and the predicted ambient temperature of the target area.

[0074] In one embodiment of the present application, the policy determination module 23 is specifically configured to: Calculate the matching degree between the waste heat prediction value and the heating demand value; When the matching degree is greater than or equal to the first matching degree threshold, a corresponding allocation strategy in the allocation strategy library is selected according to the matching degree; When the matching degree is less than the first matching degree threshold, the operating parameters of the compressor, the temperature data of the heating medium and the heating demand of the target area are processed based on the reinforcement learning algorithm to determine the allocation strategy.

[0075] In one embodiment of the present application, the policy determination module 23 is specifically configured to: Determine the initial action space vector of the reinforcement learning algorithm corresponding to the allocation strategy; Determine multiple standard allocation strategies and determine the action space vector of the reinforcement learning algorithm corresponding to each standard allocation strategy; Determine the state space vector of the reinforcement learning algorithm based on the compressor operating parameters, heating medium temperature data, and target area heating demand; determining a discount factor based on a transmission delay from the compressor to the target area; Determine the value function of the reinforcement learning algorithm based on the action space vector, state space vector, and discount factor; The allocation strategy is determined based on the initial action space vector and the feedback value obtained from the value function.

[0076] In one embodiment of the present application, the policy determination module 23 is further specifically configured to: Adjust the discount factor based on the transmission delay from the compressor to the target area; Determine the value function of the reinforcement learning algorithm based on the action space vector, state space vector, and discount factor; The discount factor is adjusted based on the transmission delay from the compressor to the target area, including: In response to the transmission delay being less than a first threshold, reducing the discount factor by a first step; In response to the transmission delay being greater than a second threshold, the discount factor is increased by a second step size, the first threshold being less than the second threshold.

[0077] In one embodiment of the present application, the policy determination module 23 is further specifically configured to: When the predicted value of waste heat is less than the heating demand value, the multi-source collaborative heating mode in the allocation strategy library is selected.

[0078] In one embodiment of the present application, a specific device controlled by a method for efficient waste heat recovery and intelligent distribution is a waste heat efficient recovery and intelligent distribution device; the device includes: Waste heat recovery module, used to recover waste heat in different temperature ranges of reciprocating natural gas compressors in stages; Heat storage and buffer module, used to store and release heat, and also to balance instantaneous heat fluctuations; Adaptive control module, which is used to analyze sensor data in real time, determine the allocation strategy, and adjust the valve opening, circulation pump speed and heat storage / release strategy according to the allocation strategy; Heating cycle module: used to switch the heat source mode according to the adaptive control module and enable the heat storage unit or other heating equipment.

[0079] In this embodiment, the waste heat recovery module is a multi-stage heat exchange device, which is respectively connected to the compressor cylinder cooling water circuit and the exhaust pipe, through a plate heat exchanger and a shell and tube heat exchanger; the waste heat recovery module of this application effectively improves the waste heat utilization efficiency through step-by-step heat exchange, and stores it in the phase change heat storage unit in combination with the set temperature to prevent the compressor load from changing and the failure shutdown from occurring, thereby ensuring the stability of heating.

[0080] The heat storage and buffer module includes: a phase change heat storage unit, which uses high thermal conductivity phase change materials (such as paraffin-based composite phase change materials) with high heat storage density and is used to store and release heat; a buffer water tank is used to be connected in series with the phase change heat storage unit to balance instantaneous heat fluctuations.

[0081] Adaptive control module, including: sensor network and heat efficient recovery and intelligent distribution system; Among them, the sensor network includes: ambient temperature sensor, indoor temperature sensor, waste heat temperature sensor, and household flow sensor.

[0082] The heating circulation module includes hot water circulation piping, connecting the thermal storage unit, a buffer water tank, and user-side radiators (such as floor heating and radiators). The heating circulation module also switches the heat source mode. When the compressor stops, it automatically activates the thermal storage unit or combines heating with the currently active gas boiler, controlling the number of gas boilers started and stopped based on temperature.

[0083] In this embodiment, the device operates as follows: When the gas reciprocating compressor is operating, the heat exchanger in the waste heat recovery module activates, transferring the waste heat generated by the compressor to the heat carrier medium. A circulating pump then pumps the heat carrier medium through pipelines to the heating terminal module, where it releases the heat to the target area. Sensors in the adaptive control module collect real-time data on ambient temperature, heat carrier temperature, pressure, and flow rate, and transmit this data to a controller within the module (the efficient heat recovery and intelligent distribution system). The controller analyzes and processes the collected data based on pre-set algorithms and strategies. For example, when the ambient temperature drops and the heating demand increases, the controller will increase the speed of the circulation pump, increase the flow of the heat carrier, and open the valve appropriately to improve the waste heat recovery efficiency to meet the indoor heating needs. If the heating effect is still not good, the currently configured gas boiler will be automatically started to ensure the heating requirements; when the ambient temperature rises and the heating demand decreases, the controller will reduce the speed of the circulation pump, reduce the flow of the heat carrier, close some valves, reduce the waste heat recovery amount, and avoid energy waste. At the same time, the heat storage unit can recycle and store excess heat to ensure continuous heating when the compressor unit fails or stops for preventive maintenance.

[0084] In this embodiment, a method for installing waste heat efficient recovery and intelligent distribution equipment is provided: Installation of waste heat recovery module: Connect the gas reciprocating compressor cylinder cooling water waste heat recovery heat exchanger in series to the compressor cylinder cooling water outlet pipeline before the compressor thermostat, and install the exhaust waste heat recovery device between the power cylinder exhaust pipe and the muffler.

[0085] Heat medium circulation module installation: Install the variable frequency circulation pump, piping, and valves, ensuring a reasonable piping layout to minimize resistance losses. The circulation pump should be installed in a location that is easily accessible for maintenance and overhaul.

[0086] Adaptive control module installation: Install temperature sensors, pressure sensors, flow sensors, etc. at key locations in the system, such as heating terminals, pipe inlets and outlets, etc. The controller should be installed in the control room or a convenient location and electrically connected to the sensors and actuators (such as circulating pumps and valves).

[0087] Heating terminal module installation: Integrate the waste heat system into the current gas boiler system to achieve automatic switching. When the waste heat is insufficient, the gas boiler can be automatically activated to ensure normal heating in winter.

[0088] The heat storage unit adopts a modular design, the capacity is set according to the heating demand, and the melting point of the phase change material is designed and selected according to the demand.

[0089] In this embodiment, the heat exchange of this application is mainly used for heat exchange utilization of the exhaust gas heat energy of the compressor unit and the jacket water heat energy, such as Figure 3 Schematic diagram of the multi-stage heat exchanger structure shown.

[0090] Specifically, the jacket water at 66-88°C is first heat exchanged through a plate heat exchanger. According to the water cooling system currently configured for the compressor unit, the AJAX DPC2803 unit can theoretically achieve a heat exchange power of 100KW. The water then enters the flue gas heat exchanger for secondary heat exchange, effectively improving the heat exchange efficiency. Theoretically, a heat exchange power of 450KW can be achieved. If only flue gas heat exchange is performed, a heat exchange power of about 350KW can be achieved according to theoretical calculations. However, due to limited load changes, the stability is low. Therefore, multi-stage heat exchange can effectively improve the heat exchange efficiency and thermal energy stability.

[0091] In one embodiment, the centralized office accommodation in the customer's operation area requires a total heating area of ​​6,822 square meters (including 3,510 square meters for dormitories, 430 square meters for offices, 1,400 square meters for comprehensive office buildings, 912 square meters for activity centers, 480 square meters for cafeterias, 50 square meters for gatehouses, and 40 square meters for fire pump rooms). Currently, three CLHG-0.35-90 / 70-AⅡ / Q gas boilers are used for heating. Each boiler has a power of 350 kW. In winter, two boilers are in operation and one is in standby mode. When the weather is extremely cold, all three boilers are started at the same time, meaning that the maximum heating power demand is 1,050 kW. The gas consumption of a single boiler is 32 m³ / h per hour. Based on a four-month heating season, the maximum gas consumption for heating in winter is 276,480 m³. The compressor unit can utilize about 900KW of waste heat energy, which is a certain difference from the customer's current heating energy of 1050KW. It can be considered to first use the compressor unit to utilize the waste heat energy for heating, and use one boiler in cold weather to ensure winter heating. According to the calculation of the coldest time of one month in Wushi Banner of Sulige Gas Field, the utilization of waste heat energy can save users 253,440m³ of natural gas each year, and achieve the beneficial effect of energy saving and environmental protection.

[0092] Through adaptive control, this application can adjust the waste heat recovery and utilization process in real time according to actual heating needs and environmental conditions, maximize the use of waste heat generated by the gas reciprocating compressor, and reduce energy waste.

[0093] See also Figure 4 , Figure 4 This is a schematic block diagram of an electronic device provided in one embodiment of the present application. Figure 4The electronic device 300 in the embodiment shown may include: one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The processors 301, input devices 302, output devices 303, and memories 304 communicate with each other via a communication bus 305. The memory 304 is used to store computer programs, which include program instructions. The processor 301 is used to execute the program instructions stored in the memory 304. The processor 301 is configured to call the program instructions to execute the functions of the modules in the above-mentioned device embodiments, such as Figure 2 The functions of the data acquisition module 21, the data prediction module 22, the data calculation module 23 and the strategy determination module 24 are shown.

[0094] It should be understood that in the embodiment of the present application, the processor 301 may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0095] The input device 302 may include a touchpad, a fingerprint collection sensor (for collecting user fingerprint information and fingerprint direction information), a microphone, etc. The output device 303 may include a display (LCD, etc.), a speaker, etc.

[0096] The memory 304 may include a read-only memory and a random access memory, and provides instructions and data to the processor 301. A portion of the memory 304 may also include a non-volatile random access memory. For example, the memory 304 may also store device type information.

[0097] In a specific implementation, the processor 301, input device 302, and output device 303 described in the embodiments of the present application can execute the implementation method described in any implementation of a method for efficient recovery and intelligent distribution of waste heat provided in the embodiments of the present application, and can also execute the implementation method of the electronic device described in the embodiments of the present application, which will not be repeated here.

[0098] In another embodiment of the present application, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program. The computer program includes program instructions. When the program instructions are executed by a processor, all or part of the process of the method in the above embodiment is implemented. The computer program can also be used to instruct related hardware to complete the process. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of each of the above method embodiments are implemented. The computer program includes computer program code, which can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium.

[0099] The computer-readable storage medium can be an internal storage unit of the electronic device in any of the aforementioned embodiments, such as a hard disk or memory of the electronic device. The computer-readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash memory card, etc. Furthermore, the computer-readable storage medium can include both an internal storage unit of the electronic device and an external storage device. The computer-readable storage medium is used to store computer programs and other programs and data required by the electronic device. The computer-readable storage medium can also be used to temporarily store data that has been output or is about to be output.

[0100] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0101] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the electronic devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0102] In the several embodiments provided in this application, it should be understood that the disclosed electronic devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces or units, or can be an electrical, mechanical or other form of connection.

[0103] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present application.

[0104] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0105] The above are only specific embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and such modifications or substitutions should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A method for efficient recovery and intelligent distribution of waste heat, characterized in that: include: Obtain the compressor's operating parameters, heating medium temperature data, and target area heating demand; Inputting the operating parameters and the heating medium temperature data into a pre-trained waste heat prediction model to generate a waste heat prediction value of recoverable waste heat; Calculating the heating demand value of the target area according to the heating demand of the target area and the thermodynamic transfer model; A corresponding allocation strategy is determined according to the waste heat prediction value and the heating demand value.

2. The waste heat efficient recovery and intelligent distribution method according to claim 1 is characterized in that: The training method of the pre-trained waste heat prediction model includes: Collect the compressor operating parameter sequence, heating medium temperature time series data and the corresponding actual waste heat recovery amount within the historical time period; The operating parameter sequence and the heating medium temperature time series data are used as input features, and the actual waste heat recovery amount is used as a label. Time series modeling training is performed through a neural network model to obtain the waste heat prediction model.

3. The waste heat efficient recovery and intelligent distribution method according to claim 1 is characterized in that: The step of calculating the heating demand value of the target area according to the heating demand of the target area and a thermodynamic transfer model includes: Calculating the heating power of the target area based on the heating demand of the target area and a thermodynamic transfer model; The heating demand includes: the ambient temperature of the target area, the building thermal resistance coefficient of the target area and the user-set temperature of the target area; The temperature fluctuation compensation amount is obtained by weighted calculation based on the historical temperature change rate of the target area and the predicted temperature difference, where the predicted temperature difference is the difference between the user-set temperature of the target area and the predicted ambient temperature of the target area.

4. The waste heat efficient recovery and intelligent distribution method according to claim 1 is characterized in that: The determining of a corresponding allocation strategy according to the waste heat prediction value and the heating demand value includes: Calculating the matching degree between the waste heat prediction value and the heating demand value; When the matching degree is greater than or equal to a first matching degree threshold, selecting a corresponding allocation strategy in an allocation strategy library according to the matching degree; When the matching degree is less than a first matching degree threshold, the operating parameters of the compressor, the temperature data of the heating medium and the heating demand of the target area are processed based on a reinforcement learning algorithm to determine an allocation strategy.

5. The waste heat efficient recovery and intelligent distribution method according to claim 4 is characterized in that: Based on the reinforcement learning algorithm, the compressor operating parameters, heating medium temperature data, and target area heating demand are processed to determine the allocation strategy, including: Determine the initial action space vector of the reinforcement learning algorithm corresponding to the allocation strategy; Determine multiple standard allocation strategies and determine the action space vector of the reinforcement learning algorithm corresponding to each standard allocation strategy; Determine the state space vector of the reinforcement learning algorithm based on the compressor operating parameters, heating medium temperature data, and target area heating demand; determining a discount factor based on a transmission delay from the compressor to the target area; determining a value function of the reinforcement learning algorithm based on the action space vector, the state space vector, and the discount factor; An allocation strategy is determined based on the initial action space vector and a feedback value obtained from the value function.

6. The waste heat efficient recovery and intelligent distribution method according to claim 5 is characterized in that: Also includes: adjusting the discount factor based on a transmission delay from the compressor to the target area; determining a value function of the reinforcement learning algorithm based on the action space vector, the state space vector, and the discount factor; The adjusting the discount factor based on the transmission delay from the compressor to the target area includes: In response to the transmission delay being less than a first threshold, reducing the discount factor by a first step; In response to the transmission delay being greater than a second threshold, the discount factor is increased by a second step size, the first threshold being smaller than the second threshold.

7. The waste heat efficient recovery and intelligent distribution method according to claim 1 is characterized in that: Also includes: When the waste heat prediction value is less than the heating demand value, the multi-source collaborative heating mode in the allocation strategy library is selected.

8. A waste heat efficient recovery and intelligent distribution system, characterized in that: include: Data acquisition module, used to obtain the operating parameters of the compressor, the temperature data of the heating medium and the heating demand of the target area; a data prediction module, configured to input the operating parameters and the heating medium temperature data into a pre-trained waste heat prediction model to generate a waste heat prediction value of recoverable waste heat; A data calculation module is used to calculate the heating demand value of the target area according to the heating demand of the target area and a thermodynamic transfer model; A strategy determination module is used to determine a corresponding allocation strategy based on the waste heat prediction value and the heating demand value.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images