Multi-mode heat supply energy consumption evaluation method and system

Abstract

The invention discloses a multi-mode heat supply energy consumption evaluation method and system, and relates to the technical field of heat supply energy consumption evaluation and dynamic thermal scheduling control, and the method comprises the steps: arranging a multi-source sensor in a high-temperature flue gas path to collect thermal parameters, and building a time sequence mapping model between a unit load and flue gas thermal characteristics; and according to the thermal power data output by the time sequence mapping model and the heat exchange characteristics of all the heat supply modes, a heat energy availability map of the target time period is generated, and heat demand adaptation data is constructed. And selecting a matching path according to energy consumption and switching constraints based on the heat energy availability map and the heat demand adaptive data, and executing mode dynamic switching control. According to the method, a complete technical closed loop from data perception to state evaluation to action execution is formed. In the multi-mode heat supply system, system fusion of three-layer logic of heat load prediction, map expression and dynamic scheduling is realized, and the intelligence, self-adaption and energy efficiency optimization capabilities of the heat supply process are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of heating energy consumption assessment and dynamic thermal scheduling control technology, specifically a multi-mode heating energy consumption assessment method and system. Background Technology

[0002] In traditional single-source heating systems, energy consumption assessment relies primarily on steady-state thermal balance theory and empirical models, which are ill-suited to complex dynamic conditions. In recent years, however, multi-energy complementarity and multi-mode coordinated heating have become a development trend in new thermal systems, typically involving parallel operation of waste heat utilization, combined heat and power (CHP), and heat pump-assisted heating. However, due to the different response inertia, regulation characteristics, and thermal efficiency levels of various modes, achieving high-precision dynamic energy consumption assessment in multi-mode heating systems remains a key technical challenge in the field of smart heating.

[0003] Existing multi-mode heating energy consumption assessment methods have significant shortcomings. At the operational condition modeling level, most assessment methods only use single-variable or linear models, which cannot accurately capture the nonlinear coupling and temporal changes between multiple source thermal parameters in the actual system. In particular, the prediction deviation is large in scenarios with drastic heat load fluctuations or alternating start-stop cycles. In terms of mode scheduling and matching, traditional methods mostly rely on static rules or threshold judgments, which are difficult to achieve mode adaptation judgment and path selection under dynamic supply and demand matching, resulting in frequent system switching or resource waste. In the response control process, existing strategies fail to effectively identify the temporal deviation relationship between thermal energy availability and user heat demand, and have not established a rollback-capable dynamic switching stability judgment mechanism, making it difficult to optimize the overall energy efficiency of the system while ensuring stable thermal output. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by this invention is that existing methods for energy consumption assessment and control of multi-mode heating systems have high linearity in operating condition modeling, insufficient accuracy in heat load prediction, lack of dynamic adaptation mechanism in mode path selection, lack of response window and stability closed-loop judgment mechanism in switching control, and the problem of how to achieve high-precision heating energy consumption assessment and dynamic mode scheduling control in a multi-mode heat source environment.

[0006] To address the aforementioned technical problems, this invention provides the following technical solution: a multi-mode heating energy consumption assessment method, comprising setting up multi-source sensors along the high-temperature flue gas path to collect thermal parameters and establishing a time-series mapping model between unit load and flue gas thermal characteristics. Based on the thermal power data output by the time-series mapping model and the heat transfer characteristics of each heating mode, a thermal energy availability map for the target time period is generated, and heat demand adaptation data is constructed. Based on the thermal energy availability map and heat demand adaptation data, a matching path is selected according to energy consumption and switching constraints, and dynamic mode switching control is executed. The process of matching the thermal power data output by the time-series mapping model with the heat transfer characteristics of each heating mode includes performing matching calculations between the thermal power data generated by the mapping model and the characteristic parameters of the heating mode corresponding to the current predicted load range. These characteristic parameters include the minimum heating start-up temperature difference, the heat exchanger thermal response inertia time, the rated heat transfer efficiency per unit time, and the upper and lower limits of the allowable heating load range. Through thermal power difference analysis, a two-dimensional matrix is ​​constructed, with the horizontal axis representing time nodes and the vertical axis representing candidate heating modes, and the matrix elements representing instantaneous supply and demand differences. Generate a thermal energy availability map, with each point in the map indicating whether the mode is compatible at the current time.

[0007] As a preferred embodiment of the multi-mode heating energy consumption assessment method described in this invention, the step of setting up multi-source sensors to collect thermal parameters in the high-temperature flue gas path includes setting up temperature acquisition devices, mass flow meters, and relative humidity sensors at the air preheater inlet, before the induced draft fan, and at the high-temperature flue gas bypass point in the flue gas duct at the tail of the boiler. A sampling period and a sliding time window are set for real-time sampling. The sampled data is associated and bound with the real-time power generation load record of the unit at the time level to form a training dataset mapping load and thermal parameters.

[0008] As a preferred embodiment of the multi-mode heating energy consumption assessment method described in this invention, the establishment of the time-series mapping model between unit load and flue gas thermal characteristics includes dividing the training dataset of load and thermal parameter mapping into power generation load intervals, constructing an independent nonlinear regression model for each load interval, with temperature, flow rate, and moisture content as input variables, and actual waste heat converted into power as the output target. The model is dynamically updated using a sliding window regression algorithm, and the prediction error and fitting residual of the model are evaluated in each sliding window cycle. If the determination coefficient of the current regression model is lower than the regression coefficient threshold, it is retrained.

[0009] As a preferred embodiment of the multi-mode heating energy consumption assessment method described in this invention, the step of constructing an independent nonlinear regression model for each load interval includes: for each power generation load interval, selecting thermal parameter sampling data within the corresponding time period as input variables, including temperature, flow rate, and moisture content, and after standardization, correlating them with the actual waste heat converted into power at the same time stamp. A nonlinear least squares method is used to construct the regression objective function, introducing quadratic and cross terms to reflect the nonlinear coupling relationship between variables, and using samples within a historical data window for parameter fitting.

[0010] As a preferred embodiment of the multi-mode heating energy consumption assessment method described in this invention, the heat power data output by the time-series mapping model and the heat exchange characteristics of each heating mode include: using real-time heat power data output by the load-heat parameter time-series mapping model, sampling at a resolution of one minute and performing structured processing; combining this with the operating status within the current power generation load range, extracting the statistical mean, fluctuation range, and maximum value of heat power. Heating mode characteristic parameters include the minimum heating start-up temperature difference, heat exchanger thermal response inertia time, rated heat exchange efficiency per unit time, and upper and lower limits of the allowable heating load range. A point-to-point heat difference comparison table is constructed between the real-time heat power data and the heating mode characteristic parameters, forming a heat supply-demand difference matrix. The horizontal axis of the matrix represents time nodes, the vertical axis represents candidate heating modes, and each element in the matrix represents the heat power difference. The heat supply-demand difference matrix is ​​normalized, and according to the heat difference threshold rule, the heating mode corresponding to each element in the matrix that satisfies the condition that the absolute value of the heat supply-demand difference is less than a preset tolerance value is marked. The preset tolerance value is no more than 10% of the heat load demand per minute. The normalized matrix and the annotation pattern are combined and encoded to generate a two-dimensional thermal energy availability map. Each point in the map marks whether the pattern is suitable at the current time.

[0011] As a preferred embodiment of the multi-mode heating energy consumption assessment method described in this invention, the step of generating a thermal energy availability map for the target time period and constructing heat demand adaptation data includes generating a mode heat load response window matrix based on a regional heat user prediction model and the required heat load at each time point within the next 15 minutes, combined with the heat exchange initiation delay, thermal inertia buffer time, and minimum operating stabilization time required for equipment switching for each heating mode. The heating and response window matrices in the thermal energy availability map are cross-filtered to retain heating modes that simultaneously satisfy the requirement of a heat power difference of less than 10% and the allowable equipment switching response time.

[0012] As a preferred embodiment of the multi-mode heating energy consumption assessment method described in this invention, the generation of the mode heat load response window matrix includes: using an LSTM prediction algorithm to predict the power generation load for each minute within the next 15 minutes; simultaneously retrieving historical heat consumption data from regional heat users and combining it with ambient temperature and time period type to construct a user heat load prediction model, outputting a user heat demand curve that matches the power generation load prediction result; setting heat response parameters for each heating mode, generating a response window interval for each heating mode at each prediction time, forming a response window matrix with time as the row and mode as the column.

[0013] As a preferred embodiment of the multi-mode heating energy consumption assessment method described in this invention, the step of selecting a matching path based on energy consumption and switching constraints includes: for each candidate heating mode path, recording its historical operational stability, number of mode switching times, and degree of adaptation to heat demand within the operating cycle. The number of mode switching times included in the candidate heating mode paths is statistically compared with the continuous adaptation time of each mode within the target cycle to filter out candidate paths with low switching frequency and long adaptation time. Under the premise of meeting heat load demand and switching window allowable conditions, the path with the highest priority is selected from the candidate paths as the heating control path for the current operating cycle.

[0014] As a preferred embodiment of the multi-mode heating energy consumption assessment method described in this invention, the dynamic switching control of the execution mode includes: real-time monitoring of the current operating mode; when the deviation between heat supply and demand exceeds 15% within three consecutive cycles and the switching response window is in a switching-allowed state, a mode switching preprocessing procedure is performed, including adjusting the heat exchanger inlet parameters, executing a slow-start command, and clearing the buffer heat storage segment. During the switching execution, the scheduler monitors whether the switching rate meets the stability condition; if abnormal temperature gradient fluctuations are detected, the switching is aborted, and the path is re-evaluated. After the switching is completed, a closed loop of one scheduling cycle is finished.

[0015] Another objective of this invention is to provide a multi-mode heating energy consumption assessment system. This system can generate a thermal energy availability map for a target time period and construct heat demand adaptation data by generating a data module that uses the thermal power data output by the time-series mapping model and the heat transfer characteristics of each heating mode. This solves the problems of existing multi-mode heating system energy consumption assessment and control methods, such as high linearity in operating condition modeling, insufficient accuracy in heat load prediction, lack of dynamic adaptation mechanism in mode path selection, lack of response window and stability closed-loop judgment mechanism in switching control, and how to achieve high-precision heating energy consumption assessment and dynamic mode scheduling control in a multi-mode heat source environment.

[0016] As a preferred embodiment of the multi-mode heating energy consumption assessment system of the present invention, it includes a model building module, a data generation and mapping module, and a dynamic mode switching module.

[0017] The model building module is used to set up multi-source sensors in the high-temperature flue gas path to collect thermal parameters and establish a time-series mapping model between unit load and flue gas thermal characteristics.

[0018] The data construction module for generating the map is used to generate a thermal energy availability map for the target time period and construct thermal demand adaptation data based on the thermal power data output by the time-series mapping model and the heat exchange characteristics of each heating mode.

[0019] The dynamic switching mode module is used to select a matching path and perform dynamic mode switching control based on the thermal energy availability map and thermal demand adaptation data, according to energy consumption and switching constraints.

[0020] A computer device includes a memory and a processor, the memory storing a computer program, the processor executing the computer program as steps to implement a multi-mode heating energy consumption assessment method.

[0021] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a multi-mode heating energy consumption assessment method.

[0022] The beneficial effects of this invention are as follows: The multi-mode heating energy consumption assessment method provided by this invention sets up multiple source sensors to collect thermal parameters in the high-temperature flue gas path, establishes a time-series mapping model between unit load and flue gas thermal characteristics, achieves high timeliness, high resolution and adaptive update of thermal power prediction capability, provides accurate data support for subsequent map construction and path judgment, significantly improves the dynamic perception capability of waste heat utilization assessment, and solves the problems of static lag and oversimplification of traditional heat balance diagrams.

[0023] Based on the heat power data output by the time-series mapping model and the heat exchange characteristics of each heating mode, a heat energy availability map for the target period is generated and heat demand adaptation data is constructed. This achieves the goal of visualizing and structurally expressing the heat source-load matching relationship, quantifying the originally fuzzy heating adaptation process into map calculation and response screening, providing graph structure input for path selection, making up for the lack of fine adaptation standards in traditional heating control, and enhancing the scientific nature and engineering feasibility of scheduling judgment.

[0024] Based on thermal energy availability maps and thermal demand adaptation data, matching paths are selected according to energy consumption and switching constraints, and dynamic switching control is implemented. This achieves the goal of improving the overall energy efficiency of the heating process and avoiding thermal shock and nonlinear instability caused by frequent switching. It significantly improves the robustness, self-regulation capability and operational reliability of the heating system, surpassing the control methods in traditional scheduling that are mainly based on manual experience or static strategies. Attached Figure Description

[0025] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 The first embodiment of the present invention provides an overall flowchart of a multi-mode heating energy consumption assessment method.

[0027] Figure 2 The following is an overall flowchart of a multi-mode heating energy consumption assessment system provided for the second embodiment of the present invention. Detailed Implementation

[0028] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0029] Example 1, referring to Figure 1 As an embodiment of the present invention, a multi-mode heating energy consumption assessment method is provided, comprising:

[0030] S1: Set up multi-source sensors in the high-temperature flue gas path to collect thermal parameters and establish a time-series mapping model between unit load and flue gas thermal characteristics.

[0031] Temperature acquisition devices, mass flow meters, and relative humidity sensors are installed at the air preheater inlet, induced draft fan inlet, and high-temperature flue gas bypass in the flue gas duct at the tail end of the boiler. Real-time sampling is performed with a sampling period of 5 seconds and a sliding time window of 10 minutes. The sampled data is correlated and bound with the real-time power generation load record of the unit at the time level to form a training dataset mapping load and thermal parameters.

[0032] A preferred scheme for the load-thermal parameter mapping training dataset specifically includes synchronizing the collected data with the real-time power generation load data of the unit recorded by the power plant central control system according to a unified timestamp format, and forming a training sample by combining the power generation load at each moment with the temperature, flow and humidity data collected at that moment, and the continuous samples constitute the thermal parameter mapping training dataset.

[0033] The training dataset mapping load and thermal parameters is divided into power generation load intervals. An independent nonlinear regression model is constructed for each load interval. The model uses temperature, flow rate, and moisture content as input variables and actual waste heat converted into power as the output target. The model is dynamically updated using a sliding window regression algorithm. The prediction error and fitting residual of the model are evaluated in each sliding window period. If the determination coefficient of the current regression model is lower than the regression coefficient threshold, it is retrained.

[0034] A preferred approach for constructing an independent nonlinear regression model for each load interval specifically includes using standardized temperature, mass flow rate, and humidity data as input variables, and the measured actual waste heat converted into power as the output variable target. A nonlinear least squares method is used for fitting. To enhance fitting accuracy, the model introduces cross and quadratic terms such as the product term between temperature and flow rate, and the square term of humidity, to reflect the potential nonlinear coupling relationships between different thermal parameters. During training, historical data windows are used as input. Within each sliding time window period, the average prediction error and the sum of squared residuals of the current model are calculated, and the coefficient of determination is used to judge the fitting effect. If the result is lower than a preset threshold of 0.85, the model for that interval is automatically retrained until the fitting accuracy meets the requirements.

[0035] It should be noted that S1 uses multi-source sensors to precisely collect flue gas thermal parameters and combines this with real-time power generation load to establish a high-precision nonlinear mapping model, enabling dynamic prediction of heat power under different load ranges. This step improves the timeliness and accuracy of waste heat assessment, providing reliable data support for subsequent multi-mode heating decisions, and possesses strong adaptability and engineering practical value.

[0036] S2: Based on the thermal power data output by the time-series mapping model and the heat exchange characteristics of each heating mode, generate a thermal energy availability map for the target time period and construct thermal demand adaptation data.

[0037] The real-time thermal power data output by the load-thermal parameter time-series mapping model is sampled at a resolution of one minute and then structured. Combined with the operating status within the current power generation load range, the statistical mean, fluctuation range and maximum value of thermal power are extracted.

[0038] A preferred scheme for extracting the statistical mean, fluctuation range, and maximum value of heat power specifically includes the following: the real-time heat power data output by the load-heat parameter time series mapping model is represented as follows:

[0039]

[0040] in, This indicates at time point t m Available thermal power calculated from the thermal power model. M represents the total number of minutes within the prediction period. m represents the current time node number.

[0041] To further extract the time-series characteristics of thermal power for difference analysis with model parameters, the statistical mean, fluctuation range, and maximum value of thermal power are defined as follows:

[0042]

[0043] in, This represents the average thermal power. Indicates the range of minimum thermal power. This indicates the range of maximum thermal power. This represents the maximum thermal power. 'w' represents the width of the historical data window used for sliding calculations.

[0044] Heating mode characteristic parameters include minimum heating start-up temperature difference Heat exchanger thermal response inertia time Rated heat exchange efficiency per unit time and the upper and lower limits of the allowable heating load range

[0045] A point-to-point heat difference comparison table is constructed between real-time heat power data and heating mode characteristic parameters to form a heat supply and demand difference matrix. The horizontal axis of the matrix represents the time node, the vertical axis represents the candidate heating mode, and each element in the matrix is ​​the heat power difference.

[0046] The heat supply and demand difference matrix is ​​normalized, and according to the heat difference threshold rule, the heating mode corresponding to each element in the matrix that satisfies the condition that the absolute value of the heat supply and demand difference is less than a preset tolerance value is marked. The preset tolerance value is no more than 10% of the heat load demand per minute. The normalized matrix and the marked mode are combined and encoded to generate a two-dimensional thermal energy availability map. Each point in the map marks whether the mode is suitable at the current time.

[0047] A preferred scheme for the heat supply-demand difference matrix specifically includes, for any time point t m and heating mode h n The theoretical heat load requirement for this model is calculated as follows:

[0048]

[0049] in, This represents the theoretical heat load demand. Indicates at t m Predicted heat load at any given time.

[0050] To enhance the ability to express different heat user types and the nonlinear time-varying characteristics of loads, a multi-input channel nested control function is combined to predict heat load. Represented as:

[0051]

[0052] Where ρ(·) represents the combination mapping of multi-factor nonlinear activation functions. U represents the number of different hot user types. κ u (·) is the local thermal behavior coupling function for user type u. User type u in τ k The time history uses nested vectors for hot behavior. This represents the environmental parameter vector of the region where this type of user is located. This indicates the energy-saving strategy impact factors corresponding to this user type (such as whether there is intelligent temperature control or whether there is a nighttime power outage). This represents the instantaneous load change rate. This is a global adjustment term that considers the mean residual of the global load forecasting model at the current moment to control output stability.

[0053] κ u (Ψ,ξ,χ) is represented as:

[0054] κ u (Ψ,ξ,χ)=γ1·||Ψ||2+γ2·log1+||ξ||1)+β3· -χ

[0055] Among them, κ u (Ψ,ξ,χ) represents the local user's heat load driving function. Ψ represents the user's historical heat consumption behavior vector over a period of time prior to the current prediction time. ξ represents the state vector of the user's external environment. χ represents factors related to the user's energy-saving strategy. γ1, γ2, and γ3 represent weighting factors.

[0056] Represented as:

[0057]

[0058] in, This indicates that the user is at τ k The load. This represents the load value at the previous moment. Δτ is the time interval.

[0059] The thermal difference can be expressed as:

[0060]

[0061] The two-dimensional thermal difference matrix is ​​represented as follows:

[0062]

[0063] Where M' represents the total number of discrete points in the time dimension of the thermal energy map, i.e., the number of minutes predicted for the future. N represents the total number of available heating modes.

[0064] Normalize all elements in the matrix:

[0065]

[0066] Normalized matrix elements If the percentage is less than 10%, then pattern h is considered valid. n At time point h m It can be adapted and correspondingly marked in the graph. The final result is a two-dimensional thermal energy availability graph, with the horizontal axis t... m , vertical axis h n The label indicates whether the content is compatible or incompatible.

[0067] Based on the regional heat user forecasting model and the required heat load at each time point in the next 15 minutes By combining the heat exchange initiation delay, thermal inertia buffer time, and minimum operating stabilization time required for equipment switching for each heating mode, a mode heat load response window matrix is ​​generated. The heating and response window matrices in the thermal energy availability map are cross-filtered to retain heating modes that simultaneously meet the requirements of a heat power difference of less than 10% and the allowable equipment switching response time.

[0068] The user heat load prediction model is expressed as follows:

[0069]

[0070] in, This represents the user heat load prediction model. This indicates the ambient temperature at the predicted time. Indicates calendar features. H past This represents historical thermal data.

[0071] For each heating mode, thermal response parameters are set. The three set parameters include start-up time, thermal response buffer duration, and minimum stable operating time. A response window interval is generated for each heating mode at each prediction time, forming a response window matrix with time as the row and mode as the column, represented as:

[0072]

[0073] Among them, R window This represents the response window matrix. This indicates the starting point of the response window, which is time point t. m The sum of startup time and startup time. This indicates the end point of the response window, which is the sum of the start point of the response window, the hot response buffer duration, and the minimum stable runtime.

[0074] It should be noted that S2 constructs a two-dimensional thermal energy availability map by multi-dimensionally matching time-series thermal power data with multi-mode heating characteristic parameters, and introduces a user heat load model and response window matrix based on LSTM prediction to achieve dynamic heat supply and demand adaptation analysis in future time periods. Its design concept is based on minute-level time-series data, embedding multi-source inputs (load behavior, environmental parameters, energy-saving strategies) into a composite function, and combining dynamic discrimination and cross-selection to ensure the timeliness and stability of mode switching. This step overcomes the limitations of existing technologies that rely on static rules or empirical thresholds for heating mode selection, and for the first time achieves a refined, quantifiable, and dynamically adjustable graphical expression of heating mode adaptability, improving the utilization efficiency of multi-mode heat sources and the level of intelligent scheduling.

[0075] S3: Based on the thermal energy availability map and thermal demand adaptation data, select the matching path according to energy consumption and switching constraints, and perform dynamic switching control of the execution mode.

[0076] For each candidate heating mode path, historical operational stability, mode switching frequency, and adaptability to heat demand are recorded within the operating cycle. The number of mode switching events within the candidate heating mode path is statistically compared with the continuous adaptation time of each mode within the target cycle to filter out candidate paths with low switching frequency and long adaptation time. The current operating mode is monitored in real time. When the deviation between heat supply and demand exceeds 15% within three consecutive cycles and the switching response window is in a switchable state, a mode switching preprocessing procedure is initiated, including adjusting heat exchanger inlet parameters, executing a slow-start command, and clearing the buffer heat storage segment. During the switching execution, the scheduler monitors whether the switching rate meets stability conditions. If abnormal temperature gradient fluctuations are detected, the switching is aborted, and the path is re-evaluated. After the switching is completed, a closed loop of one scheduling cycle is finished.

[0077] Furthermore, in this invention, low switching frequency and long adaptation time are defined as no more than 2 switching cycles within 15 minutes, and adaptation capability for more than 80% of the time period. Abnormal temperature gradient fluctuations are defined in this invention as fluctuations exceeding 5 degrees Celsius.

[0078] It should be noted that S3 constructs a mode path selection mechanism with low switching frequency and high adaptation time, and introduces heat supply and demand deviation monitoring and temperature difference fluctuation judgment to form a dynamic adaptive mode switching closed loop. This design improves the stability and response flexibility of heating scheduling, avoids energy efficiency losses and system impacts caused by frequent switching, and realizes a refined and intelligent heating control strategy.

[0079] Example 2, refer to Figure 2 As an embodiment of the present invention, a multi-mode heating energy consumption assessment system is provided, including a model building module 100, a map generation and data construction module 200, and a dynamic mode switching module 300.

[0080] S4: The model module 100 is used to set up multi-source sensors to collect thermal parameters in the high-temperature flue gas path and establish a time-series mapping model between unit load and flue gas thermal characteristics.

[0081] The model building module 100 includes a sensor deployment submodule 101 and a load thermal characteristic modeling submodule 102.

[0082] Furthermore, the sensor deployment submodule 101 is used to deploy temperature acquisition devices, mass flow meters, and relative humidity sensors at key locations along the high-temperature flue gas path at the boiler tail, specifically including the air preheater inlet, before the induced draft fan, and the high-temperature bypass pipe section. The sampling period is 5 seconds, and the sliding window is 10 minutes to ensure the acquisition of multi-source thermodynamic parameters with timeliness and stability. The load thermal characteristic modeling submodule 102 is used to construct a nonlinear mapping model between load and thermal power based on the sensor-acquired data. It adopts a sliding window regression and coefficient of determination evaluation mechanism to realize the time-series modeling of the thermal power response relationship under each power generation load range, and dynamically retrains models that do not meet the accuracy threshold.

[0083] It should be noted that the sensor deployment submodule 101 is the starting point for establishing the model module 100, determining the modeling accuracy and timeliness. The load thermal characteristic modeling submodule 102 is the foundation for subsequent spectrum construction, providing high-quality predictive support for the thermal power input data required for the spectrum.

[0084] S5: The data generation module 200 is used to generate a thermal energy availability map for the target period and construct thermal demand adaptation data based on the thermal power data output by the time-series mapping model and the heat exchange characteristics of each heating mode.

[0085] The data construction module 200 for generating maps includes a heat map generation submodule 201 and an adaptation data construction submodule 202.

[0086] Furthermore, the heat map generation submodule 201 compares and matches the predicted thermal power data with the heat exchange response characteristic parameters of different heating modes to construct a two-dimensional heat difference matrix. The horizontal axis of the heat difference matrix represents the prediction time point, and the vertical axis represents the candidate heating modes. Each element in the matrix represents the thermal adaptation difference of the mode at that time point. After normalization and heat difference threshold filtering, the matrix generates a thermal energy availability map. The adaptation data construction submodule 202, based on the load prediction curve and the user heat demand prediction model, combined with the response delay, inertia buffer, and stable operation requirements of each mode, constructs a heating mode response window matrix to identify whether each mode has switchability and adaptability in the future.

[0087] It should be noted that the heat map generation submodule 201 relies on the heat power prediction results output by S4 modeling, which is crucial for the structured representation of the heating status. The adaptation data construction submodule 202 combines the supply and demand map with the user prediction model to form a data-driven matching and judgment mechanism, which is the prerequisite for mode scheduling control.

[0088] S6: The dynamic switching mode module 300 is used to select a matching path and execute dynamic mode switching control based on the thermal energy availability map and thermal demand adaptation data, according to energy consumption and switching constraints.

[0089] The dynamic switching mode module 300 includes a path optimization submodule 301 and a switching control submodule 302.

[0090] Furthermore, the path optimization submodule 301 is used to construct a set of all feasible heating paths based on the map and response data, statistically analyze the mode switching frequency and adaptation time of each path within the prediction period, and select paths with a switching frequency of less than twice every 15 minutes and an adaptation time of more than 80% as candidate execution paths according to preset criteria. The switching control submodule 302 is responsible for monitoring the deviation of heat supply and demand in real time during operation. When the deviation exceeds 15% for three consecutive cycles and is within the switchable time period, it performs preprocessing operations such as heat exchanger inlet parameter adjustment, slow start logic, and heat buffer clearing. During the switching execution, it monitors the temperature difference gradient change. If the fluctuation is abnormal and exceeds 5°C, it immediately stops the switching and rolls back to ensure the stability and safety of the system heat exchange process.

[0091] It should be noted that the path optimization submodule 301 determines the intelligence level of the scheduling judgment. The switching control submodule 302 constructs the execution mechanism and stability guarantee logic of the scheduling response, and is the core part of the present invention to achieve multi-mode switching optimization.

[0092] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0093] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0094] More specific examples (a non-exhaustive list) of computer-readable media include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which programs can be printed, because programs can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0095] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc. It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for evaluating energy consumption in multi-mode heating, characterized in that, include: Multi-source sensors are set up in the high-temperature flue gas path to collect thermal parameters and establish a time-series mapping model between unit load and flue gas thermal characteristics. Based on the thermal power data output by the time-series mapping model and the heat exchange characteristics of each heating mode, a thermal energy availability map for the target time period is generated and thermal demand adaptation data is constructed. Based on thermal energy availability maps and thermal demand matching data, a matching path is selected according to energy consumption and switching constraints, and dynamic switching control of execution mode is implemented. The heat power data output by the time-series mapping model and the heat exchange characteristics of each heating mode include matching the heat power data generated by the mapping model with the characteristic parameters of the heating mode corresponding to the current predicted load range. The mode characteristic parameters include the minimum heating start-up temperature difference, the heat exchanger thermal response inertia time, the rated heat exchange efficiency per unit time, and the upper and lower limits of the allowable heating load range. By analyzing the difference in thermal power, a two-dimensional matrix is ​​constructed, with the horizontal axis representing time nodes and the vertical axis representing candidate heating modes. The matrix elements are the instantaneous supply and demand difference values. A thermal energy availability map is generated, with each point in the map indicating whether the mode is suitable at the current moment.

2. The multi-mode heating energy consumption assessment method as described in claim 1, characterized in that: The method of setting up multi-source sensors to collect thermal parameters in the high-temperature flue gas path includes, Temperature acquisition devices, mass flow meters, and relative humidity sensors are installed at the air preheater inlet, induced draft fan inlet, and high-temperature flue gas bypass in the flue gas duct at the tail end of the boiler. Sampling cycles and sliding time windows are set for real-time sampling. The sampled data is correlated and bound with the real-time power generation load record of the unit at the time level to form a training dataset mapping load and thermal parameters.

3. The multi-mode heating energy consumption assessment method as described in claim 1 or 2, characterized in that: The establishment of the time-series mapping model between unit load and flue gas thermal characteristics includes... The training dataset mapping load and thermal parameters is divided into power generation load intervals. An independent nonlinear regression model is constructed for each load interval. The model takes temperature, flow rate, and moisture content as input variables and actual waste heat converted into power as the output target. The model uses a sliding window regression algorithm for dynamic updates. In each sliding window period, the prediction error and fitting residual of the model are evaluated. If the coefficient of determination of the current regression model is lower than the regression coefficient threshold, it is retrained.

4. The multi-mode heating energy consumption assessment method as described in claim 3, characterized in that: The construction of an independent nonlinear regression model for each load interval includes, For each power generation load range, the thermal parameter sampling data within the corresponding time period are selected as input variables, including temperature, flow rate, and moisture content. After standardization, the data are correlated with the actual waste heat power at the same time stamp. The regression objective function is constructed using the nonlinear least squares method. Quadratic and cross terms are introduced to reflect the nonlinear coupling relationship between variables, and the parameters are fitted using samples within the historical data window.

5. The multi-mode heating energy consumption assessment method as described in claim 3, characterized in that: The heat power data output from the time-series mapping model and the heat transfer characteristics of each heating mode include, The real-time thermal power data output by the load-thermal parameter time series mapping model is sampled at a resolution of one minute and then structured. Combined with the operating status within the current power generation load range, the statistical mean, fluctuation range and maximum value of thermal power are extracted. The characteristic parameters of the heating mode include the minimum heating start-up temperature difference, the thermal response inertia time of the heat exchanger, the rated heat exchange efficiency per unit time, and the upper and lower limits of the allowable heating load range. A point-to-point heat difference comparison table is constructed between real-time heat power data and heating mode characteristic parameters to form a heat supply and demand difference matrix. The horizontal axis of the matrix represents the time node, the vertical axis represents the candidate heating mode, and each element in the matrix is ​​the heat power difference. The heat supply and demand difference matrix is ​​normalized, and according to the heat difference threshold rule, the heating mode corresponding to each element in the matrix that satisfies the absolute value of the heat supply and demand difference being less than the preset tolerance value is marked. The preset tolerance value is no more than 10% of the heat load demand per minute. The normalized matrix and the annotation pattern are combined and encoded to generate a two-dimensional thermal energy availability map. Each point in the map marks whether the pattern is suitable at the current time.

6. The multi-mode heating energy consumption assessment method as described in any one of claims 1, 2, or 5, characterized in that: The process of generating a thermal energy availability map for the target time period and constructing thermal demand adaptation data includes... Based on the regional heat user prediction model and the heat load required at each time point in the next 15 minutes, combined with the heat exchange start delay, thermal inertia buffer time, and minimum operating stabilization time required for equipment switching of each heating mode, a mode heat load response window matrix is ​​generated. Cross-filtering the heating and response window matrices in the thermal energy availability map, retaining the heating modes that simultaneously meet the requirements of a thermal power difference of less than 10% and the allowable switching response time of the equipment.

7. The multi-mode heating energy consumption assessment method as described in claim 6, characterized in that: The generated mode heat load response window matrix includes, The LSTM prediction algorithm is used to predict the power generation load for each minute within the next 15 minutes. At the same time, historical heat consumption data of regional heat users are retrieved and combined with ambient temperature and time period type to construct a user heat load prediction model and output a user heat demand curve that matches the power generation load prediction result. For each heating mode, thermal response parameters are set separately, and a response window interval is generated for each heating mode at each prediction time, forming a response window matrix with time as the row and mode as the column.

8. The multi-mode heating energy consumption assessment method as described in claim 5, characterized in that: The selection of a matching path based on energy consumption and switching constraints includes... For each candidate heating mode path, record the historical operational stability, number of mode switching times, and degree of adaptation to heat demand within the operating cycle; The number of mode switching in the candidate heating mode path is statistically compared with the continuous adaptation time of each mode in the target cycle, and candidate paths with low switching frequency and long adaptation time are selected. Provided that the heat load demand and the switching window conditions are met, the highest priority path is selected from the candidate paths as the heating control path for the current operating cycle.

9. The multi-mode heating energy consumption assessment method as described in any one of claims 1, 2, 5, or 8, characterized in that: The dynamic switching control of the execution mode includes, Real-time monitoring of the current operating mode; when the deviation between heat supply and demand exceeds 15% within three consecutive cycles and the switching response window is in the switching state, a mode switching preprocessing procedure is performed, including adjusting the heat exchanger inlet parameters, executing the slow start command, and clearing the buffer heat storage section. During the handover process, the scheduler monitors whether the handover rate meets the stability conditions. If an abnormal temperature gradient fluctuation is detected, the handover is aborted and the path is re-evaluated. Once the switchover is complete, a closed loop of one scheduling cycle is finished.

10. A multi-mode heating energy consumption assessment system, characterized in that: It includes a model building module (100), a map generation and data construction module (200), and a dynamic mode switching module (300); The model building module (100) is used to set up multi-source sensors in the high-temperature flue gas path to collect thermal parameters and establish a time-series mapping model between unit load and flue gas thermal characteristics. The data construction module (200) for generating the map is used to generate a thermal energy availability map for the target period and construct thermal demand adaptation data based on the thermal power data output by the time-series mapping model and the heat exchange characteristics of each heating mode. The dynamic switching mode module (300) is used to select a matching path based on the thermal energy availability map and thermal demand adaptation data, and to execute dynamic mode switching control according to energy consumption and switching constraints.

11. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the multi-mode heating energy consumption assessment method according to any one of claims 1 to 9.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the multi-mode heating energy consumption assessment method according to any one of claims 1 to 9.

Images