A natural gas boiler flue gas white smoke waste heat recovery energy optimization scheduling method

Abstract

This invention relates to the field of natural gas boiler energy dispatching technology, and discloses a method for optimizing energy dispatching through waste heat recovery from natural gas boiler flue gas. The method includes collecting flue gas parameters from the gas boiler and user-end heat demand parameters; determining the heat recovery threshold of the primary condensing waste heat recovery system and the heat exchange load range of the secondary white smoke elimination heat exchange system based on the flue gas parameters; generating a multi-energy coupled dispatching strategy by combining the operating status data of the four-pipe air source heat pump unit and environmental temperature and humidity parameters; allocating the heat supply ratio of the underfloor heating and air conditioning heating systems according to this strategy, and dynamically adjusting the operating modes of the two-stage systems; correcting the dispatching strategy based on real-time heat supply and demand deviation, and outputting the final energy dispatching command. This method can fully recover waste heat from the flue gas, achieve multi-energy coordinated dispatching and heat supply and demand matching, and avoid the "white smoke" phenomenon.

Description

Technical Field

[0001] This invention relates to the field of energy dispatching technology for natural gas boilers, specifically a method for optimizing energy dispatching by eliminating waste heat from flue gas in natural gas boilers. Background Technology

[0002] In the current energy utilization field, natural gas boilers are widely used as common heat supply equipment in residential heating, industrial production and other scenarios. When traditional natural gas boilers are in operation, they produce a large amount of flue gas containing waste heat. If this flue gas is directly discharged, it will not only cause serious heat loss, but may also form "white smoke" due to the water vapor carried in the flue gas, which will affect the visual effect of the surrounding environment.

[0003] While some existing technologies address waste heat recovery from boiler flue gas, most employ a single waste heat recovery system, achieving only limited heat recovery and failing to fully utilize the thermal energy in the flue gas. Furthermore, in terms of heat supply, traditional methods typically use a fixed supply ratio, separately controlling underfloor heating and air conditioning systems. This makes it difficult to flexibly adjust according to real-time heat demand from users, resulting in low energy efficiency.

[0004] Current energy dispatch strategies often overlook the impact of environmental factors, failing to effectively integrate the operating status of four-pipe air source heat pump units with ambient temperature and humidity parameters. The operating efficiency of four-pipe air source heat pump units varies under different environmental conditions. If reasonable dispatch strategies are not formulated based on their operating status and environmental parameters, energy waste will be further exacerbated. Furthermore, existing dispatch methods lack a dynamic correction mechanism for real-time heat supply and demand deviations. When user-end heat demand changes or system operation fluctuates, the dispatch plan cannot be adjusted in a timely manner, leading to a mismatch between heat supply and demand, affecting user experience, and causing additional energy consumption. These problems make it difficult for natural gas boilers to achieve optimal energy utilization during operation, which does not meet current energy conservation and emission reduction requirements. Summary of the Invention

[0005] The purpose of this invention is to provide a method for optimizing the scheduling of energy by eliminating waste heat from flue gas in natural gas boilers, in order to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides a method for optimizing energy scheduling through waste heat recovery and whitening of flue gas from natural gas boilers, the method comprising:

[0007] Collect flue gas parameters and user-end heat energy demand parameters generated during the operation of the gas-fired boiler, and determine the heat recovery threshold of the primary condensing waste heat recovery system and the heat exchange load range of the secondary whitening heat exchange system based on the flue gas parameters.

[0008] By combining the operating status data of the four-pipe air source heat pump unit with the ambient temperature and humidity parameters, a multi-energy coupling scheduling strategy is generated.

[0009] According to the multi-energy coupling scheduling strategy, the heat supply ratio of the underfloor heating system and the air conditioning heating system is allocated, and the operation mode of the primary condensing waste heat recovery system and the secondary whitening heat exchange system is dynamically adjusted.

[0010] Based on the real-time thermal energy supply and demand deviation correction multi-energy coupled scheduling strategy, the final energy scheduling command is output.

[0011] Preferably, the parameters of flue gas generated during the operation of the gas-fired boiler and the user-end heat energy demand parameters include:

[0012] Real-time monitoring of flue gas temperature, flow rate, and component concentration in gas-fired boilers generates a time-series dataset of flue gas parameters;

[0013] Acquire historical energy consumption data and real-time demand signals of user-end underfloor heating and air conditioning heating systems, and calculate the total heat energy demand per unit time.

[0014] The time series dataset of flue gas parameters and the total heat energy demand are input into the preprocessing unit. Abnormal data are removed and normalized to obtain standardized flue gas feature sequences and demand feature sequences.

[0015] Preferably, determining the heat recovery threshold of the primary condensation waste heat recovery system and the heat exchange load range of the secondary whitening elimination heat exchange system based on flue gas parameters includes:

[0016] The waste heat potential coefficient of flue gas is calculated based on the temperature and flow velocity data in the standardized flue gas characteristic sequence.

[0017] Based on the waste heat potential coefficient and the design parameters of the primary condensing waste heat recovery system, the maximum recoverable heat value and the minimum operating load boundary are derived.

[0018] Based on the characteristics of the heat exchange medium in the two-stage whitening heat exchange system and the ambient temperature and humidity parameters, the dynamic adjustment range of the heat exchange load is determined.

[0019] Preferably, the multi-energy coupled scheduling strategy includes:

[0020] Obtain refrigerant cycle status and energy efficiency ratio data for a four-pipe air source heat pump unit;

[0021] A multi-objective optimization function is constructed by integrating the heat output value of the waste heat recovery system, the real-time energy supply capacity of the heat pump unit, and the user demand characteristic sequence.

[0022] A heuristic search algorithm is used to solve the multi-objective optimization function and generate an initial set of energy scheduling strategies.

[0023] Preferably, the heat supply ratio for the underfloor heating system and the air conditioning heating system includes:

[0024] Analyze the thermal energy allocation weight coefficients in the initial energy dispatch strategy;

[0025] Based on the heating characteristics and response delay time of the underfloor heating system and the air conditioning heating system, the priority of heat energy distribution is dynamically calculated.

[0026] The system adjusts the heat supply ratio based on priority and generates a sequence of system control signals.

[0027] Preferably, the operating modes of the primary condensing waste heat recovery system and the secondary whitening heat exchange system are dynamically adjusted as follows:

[0028] Monitor the deviation between the actual recovered heat and the theoretical recovered heat of the primary condensing waste heat recovery system;

[0029] Based on the magnitude of the deviation and the real-time load status of the secondary whitening heat exchange system, switch the heat exchange mode of the secondary whitening heat exchange system.

[0030] When the actual recovered heat is consistently lower than the theoretical value, the auxiliary heating mechanism of the primary condensing waste heat recovery system is triggered.

[0031] Preferably, the multi-energy coupled scheduling strategy based on real-time thermal energy supply and demand deviation correction includes:

[0032] Collect feedback data on outlet water temperature and room temperature of underfloor heating system and air conditioning heating system;

[0033] Calculate the matching error between the actual heat supply and the demand characteristic sequence, and generate a deviation correction coefficient;

[0034] The deviation correction coefficient is fed back to the multi-objective optimization function, and the corrected energy dispatch strategy is generated by solving the problem again.

[0035] Preferably, the output final energy dispatch command includes:

[0036] The revised energy dispatch strategy is encoded into a set of device-executable instructions.

[0037] Verify the compatibility of the instruction set with the control interfaces of each system, and filter conflicting instructions;

[0038] Instructions are distributed in a time sequence to gas-fired boilers, primary condensing waste heat recovery systems, secondary whitening heat exchange systems, four-pipe air source heat pump units, underfloor heating systems, and air conditioning heating systems.

[0039] Preferably, verifying the compatibility of the instruction set with each system control interface includes:

[0040] Extract the communication protocols and data format specifications of each system;

[0041] Compare each control parameter in the instruction set with the allowable range specified in the standard.

[0042] Mark the out-of-limit instruction and replace it with a preset safe value.

[0043] Preferably, the time-series distribution instructions include:

[0044] Based on the response latency characteristics and instruction execution duration of each system, allocate instruction sending timing;

[0045] A timing synchronization mechanism is employed to ensure the coordination of instruction execution across multiple systems;

[0046] Record instruction distribution logs and update them to the energy dispatch database in real time.

[0047] Compared with the prior art, the beneficial effects of the present invention are:

[0048] By collecting flue gas parameters and user-end heat demand parameters during the operation of the gas-fired boiler, the heat recovery threshold of the primary condensing waste heat recovery system and the heat exchange load range of the secondary white smoke elimination heat exchange system are determined based on the flue gas parameters, enabling precise recovery and utilization of flue gas waste heat. Compared to traditional single waste heat recovery methods, this tiered approach to determining recovery and heat exchange parameters can more fully tap the heat potential in the flue gas, reduce heat loss caused by direct flue gas emissions, and simultaneously treat the flue gas through the secondary white smoke elimination heat exchange system to prevent the generation of "white smoke" and improve the visual environment.

[0049] This method combines the operating status data of a four-pipe air-source heat pump unit with environmental temperature and humidity parameters to generate a multi-energy coupled scheduling strategy, overcoming the limitations of traditional scheduling methods that ignore environmental factors and equipment operating status. Under different environmental temperature and humidity conditions, the operating efficiency of a four-pipe air-source heat pump unit varies. By combining its operating status with environmental parameters, this method can formulate a scheduling strategy that better reflects actual operating conditions, making the coordination between multiple energy sources more harmonious and improving overall energy utilization efficiency.

[0050] This system allocates the heat supply ratio between the underfloor heating system and the air conditioning heating system based on a multi-energy coupling scheduling strategy, and dynamically adjusts the operation modes of the primary condensing waste heat recovery system and the secondary whitening heat exchange system, changing the traditional fixed supply ratio control method. It can flexibly adjust the supply ratio of the two heating systems according to real-time changes in user heat demand, while adapting to the operation modes of the waste heat recovery and whitening heat exchange systems. This ensures that the heat supply accurately matches user needs, avoids energy waste caused by a fixed supply ratio, and improves the user experience.

[0051] A dynamic correction mechanism is constructed by correcting multi-energy coupled scheduling strategies based on real-time thermal energy supply-demand deviations and outputting final energy scheduling commands. During system operation, when user-end thermal energy demand fluctuates or equipment operation deviates, this supply-demand deviation can be promptly detected, and the scheduling strategy can be corrected accordingly. This ensures that the energy scheduling scheme always aligns with the actual supply and demand situation, avoiding additional energy consumption caused by supply-demand mismatch and further improving the rationality and economy of energy utilization. Overall, this method optimizes multiple aspects such as waste heat recovery, multi-energy coordinated scheduling, and dynamic supply-demand matching, achieving optimized energy utilization during the operation of natural gas boilers. It aligns with the development trend of energy conservation and emission reduction and has broad application prospects in residential heating and industrial thermal energy supply. Attached Figure Description

[0052] Figure 1 This is a schematic diagram illustrating the working principle of the natural gas boiler flue gas whitening and waste heat recovery energy optimization scheduling method described in this invention.

[0053] Figure 2 A flowchart for the acquisition and preprocessing of flue gas parameters and thermal energy demand parameters;

[0054] Figure 3 A flowchart generated for a multi-energy coupled scheduling strategy. Detailed Implementation

[0055] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] Please see Figure 1This invention provides a method for optimizing the scheduling of waste heat recovery energy from flue gas in a natural gas boiler. The method includes: collecting flue gas parameters generated during the operation of the gas-fired boiler, including flue gas temperature, flow rate, and component concentration; simultaneously acquiring user-end heat demand parameters, such as the real-time load of the underfloor heating system and the air conditioning heating system; based on the collected flue gas parameters, calculating and determining the heat recovery threshold of the primary condensing waste heat recovery system and the dynamic adjustment range of the heat exchange load of the secondary waste heat exchange system. Subsequently, combining the operating status data of the four-pipe air source heat pump unit and environmental temperature and humidity parameters, a multi-objective optimization function is constructed and solved using a heuristic search algorithm to generate an initial multi-energy coupling scheduling strategy. According to this strategy, the system allocates the heat supply ratio between the underfloor heating system and the air conditioning heating system, and dynamically adjusts the operating modes of the primary condensing waste heat recovery system and the secondary waste heat exchange system, for example, switching the heat exchange mode of the secondary system or triggering the auxiliary heating mechanism of the primary system based on the deviation between the actual recovered heat and the theoretical value. During operation, the system continuously monitors the deviation between the actual heat supply and user demand. By calculating the deviation correction coefficient and feeding it back to the optimization model, the scheduling strategy is corrected in real time. Finally, the corrected strategy is encoded into a set of executable instructions for the equipment. After verifying the compatibility of the instructions, they are distributed to each subsystem for execution according to the time sequence, thereby completing the optimized scheduling of energy.

[0057] Example 1: See Figure 2 Real-time monitoring of flue gas temperature, flow rate, and component concentration in gas-fired boilers is fundamental to the data acquisition phase. This monitoring is achieved through thermocouple arrays, Pitot tube flow meters, and non-dispersive infrared gas analyzers installed within the boiler flue. These sensors collect raw data once per second and package it into a time-series dataset of flue gas parameters with timestamps. Obtaining historical energy consumption data and real-time demand signals from user-end underfloor heating and air conditioning systems requires reading relevant data from the building energy management system and building automation system. Historical energy consumption data includes records of hot water flow and supply / return water temperature differences at different times of the day during past heating seasons. Real-time demand signals are derived from the difference between the room thermostat's set temperature and the actual indoor temperature, as well as feedback from the opening of electric valves on the manifold. An integrator performs time integration on the instantaneous heat load to calculate the total heat energy demand per unit time. The time-series dataset of flue gas parameters and the total heat energy demand are input into the preprocessing unit for data cleaning and standardization. The preprocessing unit uses an outlier identification algorithm based on the Laida criterion to screen the raw data stream and remove outlier data points that deviate significantly from the normal range due to instantaneous sensor drift or signal transmission loss. Then, the min-max normalization algorithm is applied to linearly transform the data with different dimensions such as temperature, flow rate, concentration and load to the range of zero to one. Finally, the standardized flue gas characteristic sequence and demand characteristic sequence with consistent scale are output.

[0058] The waste heat potential coefficient of flue gas is calculated based on temperature and flow velocity data from a standardized flue gas characteristic sequence. This calculation process comprehensively considers both the sensible heat of the flue gas and the latent heat of vaporization released by the condensation of water vapor in the flue gas. It requires referring to the flue gas component concentration data to calculate the isobaric specific heat capacity and water dew point temperature of the flue gas. Theoretical derivation is performed based on the waste heat potential coefficient and the design parameters of the primary condensation waste heat recovery system. The system design parameters include the total heat transfer area of ​​the heat exchanger, the design heat transfer coefficient, and the inlet temperature range of the cooling medium. By establishing a heat balance model, the theoretical maximum heat value that the system can recover under the current flue gas parameters can be calculated. At the same time, the minimum load boundary for safe operation of the system is determined based on the minimum wall temperature required to prevent low-temperature corrosion of the heat exchanger. By combining the characteristics of the heat exchange medium in the two-stage white plume elimination heat exchange system with the ambient temperature and humidity parameters, the dynamic adjustment range of its heat exchange load is determined. The characteristics of the heat exchange medium mainly refer to the freezing point, concentration, and specific heat capacity of antifreeze such as ethylene glycol solution. The ambient temperature and humidity parameters directly affect the diffusion conditions at the chimney outlet and the critical point for white plume formation. During the analysis, it is necessary to calculate the amount of heat that needs to be removed to cool the flue gas to saturation under different environmental conditions, so as to determine a load operating range that can effectively eliminate white plumes while avoiding excessive cooling that could lead to equipment corrosion.

[0059] The continuity and accuracy of data acquisition directly impact the reliability of the entire dispatching system. Multiple monitoring points are arranged along the flue gas flow direction to obtain the average temperature. Flow velocity measurement uses differential pressure flow meters with temperature and pressure compensation corrections. Component concentration monitoring pays particular attention to oxygen and moisture content to accurately calculate the dew point. Mining historical energy consumption data helps establish user heating patterns. Comparative analysis of data from the same period in previous years can identify different heating patterns on weekdays and holidays, and day and night. Real-time demand signal analysis requires distinguishing between the response speed and thermal inertia of underfloor heating and air conditioning systems. Underfloor heating demand relies more on the return water temperature trend, while air conditioning demand is more directly related to indoor temperature fluctuations. Robustness in the data preprocessing stage addresses unavoidable signal interference in industrial production environments. The sliding window size is adjusted based on the data sampling frequency and process dynamics. The maximum and minimum values ​​used for normalization are not fixed but rather use a dynamically updated database to record the actual range of recent data, allowing the standardization process to adapt to the slow drift of system operating conditions.

[0060] The calculation of the flue gas waste heat potential coefficient is essentially a quantitative assessment of the recoverable energy carried by the flue gas. The calculation process needs to consider that as the flue gas temperature decreases, the proportion of latent heat released when water vapor begins to condense in the total waste heat gradually increases. This makes the waste heat potential coefficient not a fixed value, but a parameter that dynamically changes with boiler load and fuel characteristics. Determining the heat recovery threshold of a primary condensing waste heat recovery system must balance maximizing energy recovery with safe equipment operation. The maximum recoverable heat value corresponds to the ideal situation of cooling the flue gas to near the temperature of the cooling medium. However, in actual operation, the limitation of acid dew point corrosion must be considered. Therefore, the setting of the minimum operating load boundary needs to reserve sufficient safety margin to ensure that the heat exchanger metal wall temperature is always higher than the acid dew point temperature of the flue gas. Determining the heat exchange load range of a secondary white smoke elimination heat exchange system is a multivariate optimization problem. Its goal is to achieve a balance between eliminating visual pollution (white smoke) and controlling energy consumption. The lower limit of the load range is determined by the dry-bulb temperature and humidity of the ambient air, requiring the flue gas to be cooled sufficiently to reduce its relative humidity. The upper limit is constrained by the economic efficiency of system pump consumption and fan energy consumption. Dynamically adjusting the range means that the range is not fixed, but is recalculated periodically or in real time with seasonal changes and weather variations.

[0061] The configuration and calibration of the sensor network are the physical foundation for this phase of work. Thermocouples need to be calibrated regularly with standard temperature sources to maintain measurement accuracy, pitot tubes need to have their pressure orifices kept clean to prevent blockage, and gas analyzers need to be calibrated by introducing standard gas as scheduled. The stability of data communication is ensured through industrial Ethernet and redundant network topology. Time-series data is augmented with packet sequence numbers and checksums during transmission, and the receiving end has a packet retransmission request mechanism to ensure the continuity of the dataset in the time dimension. The calculation model for total heat demand needs continuous feedback correction using actual heat consumption, because there is a loss between the theoretically calculated heat load and the effective heat delivered by the actual pipeline network. Regularly comparing heat meter readings with model calculations allows for correction of the loss coefficient in the calculation model. The algorithm of the preprocessing unit needs to have online learning capabilities. The threshold for identifying abnormal data should be automatically adjusted based on the historical distribution of recent data to avoid accidentally deleting valid data points during normal fluctuations in operating conditions. The normalization parameters should also be able to adaptively update as the system's operating range expands.

[0062] The coupling analysis of flue gas parameters and system design parameters is crucial for determining the recovery potential. The design parameters of the primary condensing waste heat recovery system, such as the heat exchange area, determine its theoretical maximum heat exchange capacity. However, in actual operation, the recoverable heat is always constrained by the current flue gas parameters (flow rate and temperature), and the degree of matching between the two directly affects the efficiency of waste heat recovery. Environmental parameters have a significant impact on the operating range of the secondary white smoke elimination heat exchange system. In cold, dry winters, the ambient air humidity is low, and only slight cooling of the flue gas is needed to achieve the white smoke elimination effect. At this time, the heat exchange load can be set lower to save energy. However, in the warm, humid spring and summer, the ambient air is close to saturation, requiring the flue gas to be cooled to a lower temperature to eliminate white smoke, at which point the heat exchange load demand increases significantly. The setting of a dynamic adjustment range reflects the flexibility of system operation. This range is not a fixed numerical range but a feasible region defined by multiple constraints, including equipment physical limits, chemical corrosion limitations, energy consumption economy, and environmental protection requirements. The system's control objective is to find the optimal point under the current operating conditions within this feasible region.

[0063] Example 2: See Figure 3 The acquisition of refrigerant cycle status and energy efficiency ratio (EER) data for the four-pipe air source heat pump unit forms the basis for strategy generation. Monitoring the refrigerant cycle status relies on pressure and temperature sensors installed on the high-pressure and low-pressure sides of the unit. These sensors read the refrigerant saturation temperature and pressure in real time. By comparing the theoretical cycle state parameters, the compressor's operating efficiency and any abnormalities such as blockage of the throttling mechanism can be determined. The EER data is directly provided by the unit's built-in energy management module, which integrates the real-time power and heating capacity calculations of all energy-consuming components such as the compressor, fan, and water pump. Integrating the output heat value of the waste heat recovery system, the real-time energy supply capacity of the heat pump unit, and the user-end demand characteristic sequence is the core step in building the optimization model. The output heat of the waste heat recovery system is calculated using heat meters installed on the primary and secondary sides. The real-time energy supply capacity of the heat pump unit is dynamically assessed based on its current operating frequency, evaporation temperature, and condensation temperature. The user-end demand characteristic sequence reflects the hourly heat load forecast within a future scheduling cycle. These three types of data are synchronized to the central dispatcher's data buffer.

[0064] Constructing a multi-objective optimization function requires clearly defining the system's optimization objectives and operational constraints. In this implementation, the optimization objectives primarily include minimizing the total system power consumption, maximizing the waste heat recovery from the natural gas boiler flue gas, and minimizing the comfort deviation of the user's room temperature from the set value. Operational constraints cover equipment physical limits such as the maximum and minimum heating capacity of the heat pump unit and the safe flow range of the waste heat recovery system, as well as system stability requirements such as boiler load change rate limits and the upper limit of the number of start-ups and shutdowns of the heat pump unit per unit time. Solving this multi-objective optimization function using a heuristic search algorithm involves a complex iterative calculation process. The preferred algorithm is an improved multi-objective particle swarm optimization algorithm. In the initialization phase, this algorithm randomly generates a group of particles in the solution space. The position vector of each particle encodes the set values ​​of all controllable variables, such as the boiler load rate, heat pump compressor frequency, and waste heat recovery system pump frequency, at each time step within the scheduling cycle. The particle's velocity vector determines its search direction and step size in the solution space.

[0065] The iterative process of the Particle Swarm Optimization (PSO) algorithm guides the search direction by continuously updating the individual and swarm historical best positions of particles. Each particle's fitness value is calculated by decoding its position vector into a specific equipment operation plan, which is then substituted into the aforementioned multi-objective function. Since conflicts exist between objective functions—for example, pursuing minimum power consumption might lead to insufficient waste heat recovery—the algorithm seeks a set of non-dominated solutions, i.e., Pareto optimal solutions. Generating an initial set of energy scheduling strategies is a direct result of the algorithm's output. Each solution in the Pareto solution set represents a feasible energy scheduling plan, with different trade-offs among multiple optimization objectives. The scheduling system needs to select a final plan from the solution set according to a set of preset decision rules. These rules can be weighted summation of different objectives or a primary objective directly specified based on the current operating priority, selecting the optimal solution for that objective.

[0066] The accuracy of operating status data for four-pipe heat pump units is crucial. Pressure sensors must meet industrial-grade accuracy standards and be calibrated regularly to prevent misjudgments of refrigerant status due to sensor drift. Calculating the energy efficiency ratio requires accurate instantaneous electrical and heating power data from the unit controller; any data transmission delay or loss will affect the real-time performance of optimization calculations. The fusion of multi-source data faces the challenge of time synchronization. Data acquisition timestamps from the boiler, heat pump, and user sides must be aligned using a clock synchronization protocol; otherwise, optimization calculations based on misaligned data will produce deviations. The central dispatcher has a data validity check module that discards data packets exceeding a reasonable time error range and requests retransmission. Setting constraints for optimization functions requires sufficient engineering experience. Equipment physical limits can be obtained from equipment technical manuals, but system stability constraints, such as load change rate limits, need to be determined through experimental or historical operating data analysis based on the thermal inertia of the actual pipe network. Overly conservative constraints will sacrifice system adjustment flexibility, while overly lenient constraints may cause system oscillations.

[0067] The parameter configuration of heuristic algorithms has a significant impact on solution efficiency and quality. In particle swarm optimization, the inertia weight determines the tendency of particles to maintain their original flight speed; larger weights are beneficial for global search, while smaller weights are beneficial for local fine-grained search. The learning factor controls the acceleration of particles moving towards their individual and swarm optimal positions. These parameters typically require pre-simulation debugging for specific energy system scales. The presentation of the Pareto optimal solution set needs to be easy for operators to understand. Advanced graphical interfaces typically project multidimensional objective functions onto two- or three-dimensional space for visualization, helping scheduling engineers intuitively grasp the balance between energy consumption, waste heat recovery, and comfort among different scheduling schemes. The final scheme selection sometimes also requires fine-tuning based on human experience. The generation of the initial energy scheduling strategy set is periodically triggered. At the beginning of each scheduling cycle, the system re-executes a complete optimization calculation to cope with the latest changes in external environmental parameters and user needs. However, the system also has the ability to perform rolling optimization; if a major disturbance is detected in the middle of the cycle, the current plan can be interrupted and recalculated.

[0068] In-depth analysis of refrigerant cycle status helps predict the performance degradation of heat pump units. For example, long-term tracking of the compressor's compression ratio and isentropic efficiency can determine its internal wear. Although this information is not directly involved in the current optimization calculation, it can provide data support for preventative maintenance, indirectly ensuring the reliability of the long-term execution of the optimization strategy. Multi-objective optimization models implicitly include economic indicators. Although cost terms are not directly present in the objective function, minimizing electricity and gas consumption essentially corresponds to minimizing operating costs, while maximizing waste heat recovery improves energy utilization efficiency, equivalent to reducing the unit heating cost. Comfort deviations are related to the potential hidden gains or losses in service quality. Particle swarm optimization (PSO) exhibits good robustness in solving such high-dimensional, nonlinear, and strongly constrained optimization problems. Its swarm intelligence characteristics make it less prone to getting trapped in local optima, making it particularly suitable for handling engineering problems with many variables and complex coupling relationships, such as energy scheduling. Its computation time also meets the real-time requirements of most practical applications.

[0069] The generation of the initial strategy set does not signify the end of the optimization process, but rather marks the beginning of a dynamic scheduling loop. The generated strategies need to retain a certain degree of redundancy and flexibility to cope with the unavoidable uncertainties in real-time operation. For example, the strategy should set a reasonable load adjustment range for the heat pump unit rather than a fixed value, allowing the downstream real-time control loop to fine-tune according to instantaneous supply and demand fluctuations. The entire strategy generation process places certain demands on the performance of the computing hardware. Complex multi-objective optimization algorithms need to complete calculations within minutes to issue instructions promptly. Therefore, the central scheduler is typically equipped with high-performance industrial servers and employs parallel computing technology to accelerate the iteration process. The algorithm code also needs to be fully optimized to improve computational efficiency.

[0070] Example 3: Analyzing the heat energy allocation weight coefficients in the initial energy dispatch strategy is the primary task in the execution phase. These weight coefficients typically exist as a vector, where each element corresponds to the allocation ratio of the underfloor heating system and the air conditioning heating system in the total heating load over a specific time period. The analysis process requires extracting dimensional data related to load allocation from the solution vector output by the optimization algorithm. Based on the heating characteristics and response delay times of the underfloor heating and air conditioning heating systems, the priority of heat energy allocation is dynamically calculated. The underfloor heating system exhibits significant thermal inertia, radiating heat through heating the concrete floor slab or infill layer. It takes a considerable amount of time for a significant change in room temperature to occur after an adjustment command is issued. In contrast, air conditioning heating systems typically use fan coil units for forced convection heat exchange, enabling rapid response to load changes but relatively weaker ability to maintain stable room temperature. The response delay time is obtained by fitting historical operating data; the delay time constant of the underfloor heating system is much larger than that of the air conditioning system. Adjusting the heat supply ratio based on calculated priorities involves complex real-time decision-making. When the system predicts a significant increase in heat load in the near future, such as when user activity begins to increase in the early morning, it is necessary to increase the allocation ratio of the air conditioning system in advance to utilize its rapid response capability to prevent room temperature drop. Conversely, when the load tends to stabilize or decrease slowly, the proportion of the underfloor heating system, which has good temperature maintenance capabilities, should be increased to achieve more stable and energy-efficient operation.

[0071] Generating the system control signal sequence is a crucial step in translating the allocation strategy into specific equipment instructions. This signal sequence contains control commands for different actuators, such as valve opening instructions sent to the electric regulating valves on the underfloor heating system manifold, and speed setpoint instructions sent to the air conditioning unit fan inverter. Each command in the instruction sequence has a precise execution timestamp to ensure correct timing. Monitoring the deviation between the actual and theoretical recovered heat of the primary condensing waste heat recovery system is the basis for adjusting the operating mode. The actual recovered heat is monitored through temperature sensors and flow meters installed at the inlet and outlet of the working fluid in the waste heat recovery system. The measured value is obtained by calculating the heat absorbed by the working fluid per unit time, while the theoretical recovered heat is calculated based on the current flue gas parameters and the system design model. The decision to switch the heat exchange mode of the secondary whitening heat exchange system is based on the magnitude of the deviation and the real-time load status of the secondary whitening heat exchange system. The secondary whitening heat exchange system is typically designed with multiple operating modes, such as single-stage heat exchange mode, two-stage series heat exchange mode, or bypass mode. Its real-time load status is assessed by monitoring the inlet and outlet temperatures and flow rates of its heat exchange medium.

[0072] When the actual recovered heat remains consistently lower than the theoretical value and the deviation exceeds a set threshold for a period of time, the system will trigger the preset auxiliary heating mechanism of the primary condensing waste heat recovery system. This mechanism may include activating the electric auxiliary heater built downstream of the heat exchanger to supplement the heating of the flue gas, or opening a bypass valve connected to the high-temperature flue gas of the boiler to introduce a small amount of high-temperature flue gas to raise the temperature of the mixed flue gas. The aim is to raise the flue gas temperature to a level more conducive to water vapor condensation, thereby improving the waste heat recovery effect. The dynamic calculation process of heat energy allocation priority can be achieved using an evaluation function that comprehensively considers the system's response requirements and operating energy efficiency.

[0073]

[0074] Wherein: P represents the calculation priority score of a heating system at the current moment; a higher score indicates that it should be given priority in heat allocation. τ represents the characteristic response time required for the heating system to significantly impact room temperature from receiving control commands; this time is obtained by analyzing the system's historical step response data. η represents the overall operating efficiency of the heating system at the current expected operating load point; this efficiency value considers the energy consumption of its driving equipment (such as pumps and fans) and the effective heat output. α and β are two weighting coefficients used to adjust the relative importance of response speed and operating efficiency in priority evaluation. The specific values ​​of these coefficients need to be adjusted according to the overall operating strategy of the entire heating system. For example, a larger value for α is assigned during periods requiring rapid tracking of load changes, while a larger value for β is assigned during periods emphasizing the overall economic operation of the system.

[0075] The generation of control signal sequences needs to be compatible with the communication protocol of the building automation system. Standard protocols such as BACnet or Modbus are typically used to convert floating-point setpoints into corresponding register values. Range checks are performed before instructions are issued to prevent exceeding the actuator's operating limits. Deviation monitoring is continuous. The system calculates the average deviation within a sliding time window to eliminate the impact of instantaneous fluctuations and avoid malfunctions caused by a single abnormal data point. Mode switching logic is only activated when the deviation persists and shows a clear trend. Mode switching in the secondary white heat exchange system involves the coordinated operation of multiple valves. The control logic must ensure a smooth and orderly switching process to avoid pressure shocks to the flue gas system. A smooth switching method, typically opening the new path first and then closing the old one, is usually adopted. The trigger conditions for the auxiliary heating mechanism need careful setting. It must ensure timely compensation when waste heat recovery is insufficient while preventing frequent start-ups and shutdowns of auxiliary equipment from reducing its lifespan. Therefore, in addition to the deviation threshold, a minimum continuous trigger time is usually set. Auxiliary heating will only be activated if the deviation exceeds the threshold and continues for that time.

[0076] Determining the response delay time is not a one-time process; it drifts slowly with network characteristics, equipment aging, and even seasonal changes. Therefore, the system incorporates a self-learning algorithm that periodically updates the parameter values ​​in the delay time database by comparing the historical command issuance time with the corresponding room temperature change start time. Adjusting the heat energy distribution ratio is a closed-loop process. After issuing a control command, the system continuously monitors the actual changes in the underfloor heating water supply temperature and the air conditioning supply temperature. If the actual change trajectory deviates significantly from the expected model, the subsequent distribution weight coefficients are fine-tuned. This online correction capability enhances the system's adaptability to nonlinear and time-varying systems. Deviation analysis of the actual recovered heat in the primary system helps diagnose the system status. If the deviation is consistently negative and gradually increases, it may indicate that scaling on the outer wall of the heat exchanger finned tubes is causing a decrease in heat transfer efficiency. In this case, in addition to triggering auxiliary heating, the system should also send a cleaning reminder to maintenance personnel, achieving predictive maintenance functionality.

[0077] The switching strategy for the operating mode of the secondary white smoke elimination heat exchange system needs to be closely coupled with environmental conditions. For example, when the ambient temperature is extremely low and the humidity is high, the heat exchange required to eliminate white smoke is very large. In this case, even if the primary system recovers well, the secondary system may need to operate in a high-load two-stage series mode to enhance the cooling effect. When the ambient temperature is high, it may switch to a single-stage mode or even a partial bypass mode to save water pump power consumption. The degree of intervention of the auxiliary heating mechanism needs to be optimized and controlled. Its heating power is not simply set to the maximum value, but is proportionally adjusted according to the difference between the actual recovered heat and the target value, striving to make up for the gap in waste heat recovery with minimal auxiliary energy consumption and avoid "overcompensation" that causes new energy waste. The entire dynamic adjustment process reflects the complexity of multi-system coordinated control. The distribution adjustment of the underfloor heating and air conditioning systems will change the return water temperature of the entire heating network, which in turn affects the condensing temperature and recovery efficiency of the primary waste heat recovery system. The operating mode of the secondary white smoke elimination system directly affects the exhaust state of the chimney. Therefore, the issuance of these adjustment commands needs to be finely coordinated in terms of time to ensure that the system smoothly transitions from one steady state to another.

[0078] Example 4: Collecting outlet water temperature and room temperature feedback data from the underfloor heating and air conditioning heating systems forms the sensing basis for deviation correction. Taking an office building using this system as an example, the outlet water temperature monitoring point of the underfloor heating system is located on the main water supply pipe of the manifold, measured using a platinum resistance temperature sensor. The outlet water temperature of the air conditioning heating system refers to the supply water temperature of the fan coil unit. Room temperature feedback data is obtained through an indoor temperature and humidity sensor network installed in various main functional areas (such as open-plan offices, private offices, and meeting rooms). These sensors upload data to the building automation system at a one-minute interval. Calculating the matching error between the actual heat supply and the demand characteristic sequence first requires converting the collected physical quantities into heat values. The actual heat supply of the underfloor heating system is calculated using the formula "Heat = Flow Rate × Specific Heat Capacity × Supply and Return Water Temperature Difference," where the flow rate is measured by an electromagnetic flowmeter. The calculation of the heat supply of the air conditioning system is similar but requires consideration of air-side heat exchange. The demand characteristic sequence comes from the hourly heat load values ​​generated by the short-term load forecasting model described in the implementation method over a future period. The matching error is defined as the difference between the actual heat supply and the predicted demand at the same time point.

[0079] Generating deviation correction coefficients is a crucial step in transforming physical error quantities into control parameters. The system doesn't directly use the original error values; instead, it calculates the integral of the error over a sliding time window and its changing trend. For example, the system analyzes error data from the past six time points, calculates its weighted average and slope, and generates a proportional correction coefficient suitable for the next scheduling cycle. This coefficient is used to adjust the weights of corresponding terms in the objective function. Feeding the deviation correction coefficients back to the multi-objective optimization function means adjusting the internal parameters of the optimization problem online. For instance, if the actual heat supply exceeds the predicted demand for several consecutive cycles, it indicates that the prediction model may be conservative. In this case, the generated correction coefficients adjust the energy-related weights in the function, making the new solution more inclined to reduce equipment output to achieve energy conservation. The optimization algorithm then iterates again after incorporating the new weights.

[0080] Generating a revised energy dispatch strategy is the core of closed-loop control. The re-solved strategy will reflect the correction of the operational deviations of the previous cycle. For example, the new strategy might suggest appropriately reducing the initial set load of the gas boiler or delaying the start-up time of the air source heat pump unit. The amount of strategy correction is proportional to the magnitude of the calculated deviation correction coefficient. Encoding the revised energy dispatch strategy into a set of executable instructions requires strict adherence to protocol specifications. The instruction set typically exists in the form of a structured data table. Each instruction includes the device address code, function code, register start address, data length, and specific set value data. For example, an instruction sent to the boiler controller might adjust the load setpoint from 75% to 70%.

[0081] Verifying the compatibility of the instruction set with the control interfaces of various systems is a crucial safety measure to prevent misoperation. The verification process first checks whether the device address code in the instruction is in the system's configured device address list, then verifies whether the function code is supported by the target device, and finally checks whether the set value data exceeds the upper and lower limits allowed by the device parameters. Any instruction found to be incompatible during the verification process will be marked and its issuance will be suspended. Filtering conflicting instructions relies on a predefined table of device interlocking logic relationships. For example, when the instruction set contains both "start the gas boiler" and "close the gas valve at the boiler front end," the system will identify the logical conflict and filter out the latter according to priority rules to ensure the logical consistency of the instructions. Refer to Table 1, which shows a simplified instruction set verification process, listing some of the instructions to be issued and their verification results.

[0082] Table 1: Verification of Energy Dispatch Instruction Set

[0083]

[0084] The spatial representativeness of room temperature feedback data needs careful consideration. For large open office areas, a single temperature measurement point may not reflect the temperature of the entire area. Therefore, multiple sensors are usually deployed, and their average or highest value is taken as the representative temperature of the area to prevent system misjudgment due to local temperature distortion. The accuracy of the actual heat supply calculation depends on the accuracy of flow and temperature measurements. Electromagnetic flowmeters need to be zero-point calibrated regularly to eliminate drift. The installation position of platinum resistance temperature sensors must ensure sufficient straight pipe length to avoid inaccurate temperature measurement due to eddy currents. These measurement errors will be directly transmitted to the calculation of matching error. The algorithm for generating deviation correction coefficients needs to be anti-interference capable to avoid overreacting to short-term, random fluctuations. The algorithm usually sets a dead zone. The coefficient calculation is only triggered when the error continuously exceeds the dead zone range and remains there for a certain period of time, which ensures the stability of system control. Online adjustment of the optimization function weights reflects the system's learning and adaptive capabilities. The adjustment process is not entirely autonomous. System administrators can set limits on the magnitude of weight adjustments to prevent system oscillation caused by excessive single corrections. They can also lock certain weights under certain operating modes to maintain the consistency of the control strategy. The computational load of resolving and generating the corrected strategy needs to be considered. Since it runs online in real-time, the computation time of the optimization algorithm must be strictly limited. Therefore, sometimes a simplified model or a strategy of local search based on the previous cycle's optimal solution is used, sacrificing some theoretical optimality for real-time computation. The reliability of instruction encoding depends on the complete implementation of the communication protocol stack. From application-layer data encapsulation to physical-layer signal transmission, each link requires error detection and correction mechanisms. Industrial Ethernet switches typically use a ring network topology to ensure that instructions can still be delivered via a backup path if one communication path is interrupted.

[0085] Instruction set compatibility verification can be viewed as a security firewall, intercepting potentially dangerous instructions caused by model calculation errors or communication interference. For example, a huge load setpoint resulting from data overflow will be truncated to the maximum allowed value during the verification phase, thus avoiding the risk of equipment overload. The filtering logic for conflicting instructions is based on a deep understanding of the system's processes. These interlocking logics are typically stored in the system as a rule base. This rule base needs to be updated and maintained as equipment hardware changes or process optimizations occur. For instance, adding a new water pump requires adding its related start-stop interlocking logic to the rule base. The final issued instruction set is a safe and reliable set after multiple verifications. These instructions are distributed to the field controllers according to a preset time sequence, driving the entire heating system towards the corrected optimization target, thus completing a full closed-loop control cycle of monitoring, calculation, correction, and execution.

[0086] Example 5: Extracting the communication protocols and data format specifications of each system is a prerequisite step for performing instruction compatibility verification. This process is completed through a centralized protocol library, which stores the technical documents of the control interfaces of all connected systems. For example, a gas boiler controller may use the Modbus RTU protocol, whose data format specification stipulates that the load setpoint should be written to the holding register address 40001, the data type is a 16-bit unsigned integer, and the range 0-10000 corresponds to 0%-100% load. A four-pipe air source heat pump unit may use the BACnetIP protocol, and its start-stop control is implemented by writing to the Present_Value attribute of the binary output object, while the operating frequency setpoint is written to the analog output object, with a range of 0-50Hz corresponding to 0-100% frequency. The process of comparing each control parameter in the instruction set with the allowable range of the specification is an automated verification process. The system matches the queue of instructions to be issued with the specifications in the protocol library. For example, if an instruction sent to the boiler controller is "write address 40001, value 8000", the verification program will check whether address 40001 exists and confirm that the value 8000 is within the valid range of 0-10000. For the instruction of the heat pump unit "write analog output object AV:10, value 80.5", it will check whether object AV:10 is writable and confirm that the value 80.5 is within the set range of 0-100.0.

[0087] Marking out-of-limit commands and replacing them with preset safe values ​​is a fail-safe handling mechanism. When the parameters of a command exceed the allowable range, the system does not simply discard the command, but marks the anomaly in the alarm log and corrects the command parameters to preset safe values. For example, if a command attempts to set the boiler load to 110%, while the allowable upper limit is 100%, the system will mark the command as "out of limit" and replace the set value with the safe upper limit of 100%. Similarly, if a command requires the heat pump to operate at -10Hz, it will be corrected to the safe lower limit of 0Hz. The timing of command transmission needs precise time planning based on the response latency characteristics and command execution duration of each system. Response latency characteristics are obtained through preliminary testing. For example, a gas boiler may need two to three minutes from receiving a load command to the flue gas parameters stabilizing, while the entire opening and closing action of an electric regulating valve may only take thirty seconds. Command execution duration refers to the time required for the command to be processed in the communication network and within the equipment. The goal of timing allocation is to ensure that the actions of all devices are ultimately coordinated and consistent on the timeline to achieve the predetermined system state. The timing synchronization mechanism to ensure the coordinated execution of instructions from multiple systems typically relies on a high-precision network time protocol. The scheduling server and all field controllers are synchronized to the same time source with millisecond-level accuracy. Each instruction is assigned a precise absolute execution timestamp, such as "issue the boiler instruction at T0+10 seconds and the heat pump instruction at T0+12 seconds." After receiving the instruction, the field controller caches it until the time specified by its timestamp before execution. This mechanism effectively compensates for the uncertainty of network transmission time. Recording instruction distribution logs and updating them in real time to the energy scheduling database forms the basis for system operation auditing. The logs record at least the instruction content, target device identifier, planned execution timestamp, actual issuance time, and execution result feedback status (such as success, failure, timeout). These logs are not only used for post-event traceability and analysis but also provide data support for optimizing instruction timing and evaluating system response performance.

[0088] The establishment and maintenance of the protocol library is an ongoing process. Whenever a new device is connected or the firmware of an existing device is upgraded, the corresponding entries in the protocol library need to be updated. This includes communication port parameters, baud rate, data bits, stop bits, parity check, and detailed definitions of all registers or object points used for control. Even a minor parameter definition error can lead to control failure. The depth of instruction comparison goes beyond numerical range checks to include semantic checks. For example, for boiler controllers, it checks whether the written register address is indeed a control parameter address and not a read-only status address. For devices using complex protocols such as BACnet, it is also necessary to check whether the type of the written object is correct, avoiding writing a digital value to an analog object. Setting preset safety values ​​requires deep process knowledge. Safety values ​​are not always the limits of equipment operation; sometimes they may be values ​​maintained for the current state. For example, when sending an over-limit frequency command to a running water pump, a safer choice might be to maintain its setpoint at the current frequency rather than forcibly setting it to the upper or lower limit, avoiding drastic fluctuations in pipeline pressure.

[0089] The quantification of response latency characteristics is uncertain. The response latency of the same device may vary slightly depending on its operating years, maintenance conditions, and even seasonal changes. Therefore, systems sometimes introduce adaptive adjustment mechanisms, dynamically fine-tuning the lead time of commands by analyzing the time difference between the issuance time of historical commands and the time it takes for the device to reach a steady state. Timing allocation schemes need to consider the logical dependencies between commands. For example, when starting a heat exchange system, the circulating water pump usually needs to be started first. After confirming that the pump is operating normally, the steam or hot water valve is then opened. For commands with a strict sequence, the timestamp setting must include sufficient safety intervals to ensure that the preceding actions have been completed. The stability of network time synchronization is crucial. In industrial environments, hard-wired methods are sometimes used to transmit synchronization signals as a backup for the network time protocol, ensuring that the system can maintain basic time synchronization order and prevent command execution chaos when the master clock signal is lost.

[0090] Managing instruction distribution logs requires balancing detail and storage space. Overly detailed logs quickly consume large amounts of storage, while overly brief logs hinder fault analysis. Typically, systems employ a tiered logging strategy: summaries are recorded for normal operations, while complete data packets are recorded for abnormal events to facilitate in-depth analysis. Real-time database updates mean that the status of instruction issuance needs to be refreshed frequently on the human-machine interface, allowing operators to monitor the entire system's execution status almost in real time. The database design needs to optimize read and write performance to handle high-frequency concurrent access. The entire instruction verification and distribution process constitutes the final execution checkpoint of the control system. Its reliability directly determines whether the optimized scheduling strategy can be accurately implemented in the physical world. Any oversight in any link may render the meticulous calculations made earlier meaningless, or even cause the system to deviate from its safe operating range.

[0091] The logic for replacing out-of-limit instructions needs to be scenario-aware. For example, during a specific phase of system startup, a parameter briefly exceeding its limit might be normal. Directly replacing it with a safe value in this case could hinder the startup process. Therefore, the system allows defining different safe value replacement strategies for different operating conditions, increasing control flexibility. The timing synchronization mechanism demonstrates robustness in dealing with network jitter. Since instructions are executed with timestamps and caches, as long as network latency fluctuates within a specified window, instructions can be executed at precise times. This overcomes the control asynchrony problem caused by uncertain network latency in traditional request-response models. The authenticity of log records needs to be guaranteed. The system typically uses timestamped digital signature technology to protect critical instruction logs, preventing malicious tampering and providing credible evidence for incident investigations. Finally, after all instructions have undergone rigorous verification, been distributed in precise timing, and been fully recorded, a complete control loop from policy generation to physical execution is considered complete, and the system's energy optimization goals are gradually achieved in a real-world environment.

[0092] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0093] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for optimizing energy scheduling through waste heat recovery and whitening of flue gas from natural gas boilers, characterized in that: The method includes: Collect flue gas parameters and user-end heat energy demand parameters generated during the operation of the gas-fired boiler, and determine the heat recovery threshold of the primary condensing waste heat recovery system and the heat exchange load range of the secondary whitening heat exchange system based on the flue gas parameters. By combining the operating status data of the four-pipe air source heat pump unit with the ambient temperature and humidity parameters, a multi-energy coupling scheduling strategy is generated. According to the multi-energy coupling scheduling strategy, the heat supply ratio of the underfloor heating system and the air conditioning heating system is allocated, and the operation mode of the primary condensing waste heat recovery system and the secondary whitening heat exchange system is dynamically adjusted. Based on the real-time thermal energy supply and demand deviation correction multi-energy coupled scheduling strategy, the final energy scheduling command is output. The parameters of flue gas generated during the operation of the gas-fired boiler and the user-end heat energy demand parameters collected include: Real-time monitoring of flue gas temperature, flow rate, and component concentration in gas-fired boilers generates a time-series dataset of flue gas parameters; Acquire historical energy consumption data and real-time demand signals of user-end underfloor heating and air conditioning heating systems, and calculate the total heat energy demand per unit time. The flue gas parameter time series dataset and the total heat energy demand are input into the preprocessing unit, outlier data are removed and normalized to obtain standardized flue gas feature sequences and demand feature sequences. The heat recovery threshold of the primary condensing waste heat recovery system and the heat exchange load range of the secondary whitening elimination heat exchange system are determined based on flue gas parameters, including: The waste heat potential coefficient of flue gas is calculated based on the temperature and flow velocity data in the standardized flue gas characteristic sequence. Based on the waste heat potential coefficient and the design parameters of the primary condensing waste heat recovery system, the maximum recoverable heat value and the minimum operating load boundary are derived. Based on the characteristics of the heat exchange medium in the two-stage whitening heat exchange system and the ambient temperature and humidity parameters, the dynamic adjustment range of the heat exchange load is determined. The generation of multi-energy coupled scheduling strategies includes: Obtain refrigerant cycle status and energy efficiency ratio data for a four-pipe air source heat pump unit; A multi-objective optimization function is constructed by integrating the heat output value of the waste heat recovery system, the real-time energy supply capacity of the heat pump unit, and the user demand characteristic sequence. A heuristic search algorithm is used to solve the multi-objective optimization function and generate an initial set of energy scheduling strategies.

2. The method for optimizing energy scheduling of natural gas boiler flue gas waste heat recovery and elimination as described in claim 1, characterized in that, The heat supply ratio for the underfloor heating system and the air conditioning heating system includes: Analyze the thermal energy allocation weight coefficients in the initial energy dispatch strategy; Based on the heating characteristics and response delay time of the underfloor heating system and the air conditioning heating system, the priority of heat energy distribution is dynamically calculated. The system adjusts the heat supply ratio based on priority and generates a sequence of system control signals.

3. The method for optimizing energy scheduling of natural gas boiler flue gas whitening and waste heat recovery according to claim 2, characterized in that, Dynamically adjusting the operating modes of the primary condensing waste heat recovery system and the secondary whitening heat exchange system includes: Monitor the deviation between the actual recovered heat and the theoretical recovered heat of the primary condensing waste heat recovery system; Based on the magnitude of the deviation and the real-time load status of the secondary whitening heat exchange system, switch the heat exchange mode of the secondary whitening heat exchange system. When the actual recovered heat is consistently lower than the theoretical value, the auxiliary heating mechanism of the primary condensing waste heat recovery system is triggered.

4. The method for optimizing energy scheduling of natural gas boiler flue gas whitening and waste heat recovery according to claim 3, characterized in that, Multi-energy coupled scheduling strategies based on real-time thermal energy supply-demand deviation correction include: Collect feedback data on outlet water temperature and room temperature of underfloor heating system and air conditioning heating system; Calculate the matching error between the actual heat supply and the demand characteristic sequence, and generate a deviation correction coefficient; The deviation correction coefficient is fed back to the multi-objective optimization function, and the corrected energy dispatch strategy is generated by solving the problem again.

5. The method for optimizing energy scheduling of natural gas boiler flue gas whitening and waste heat recovery according to claim 4, characterized in that, The final energy dispatch instructions output include: The revised energy dispatch strategy is encoded into a set of device-executable instructions. Verify the compatibility of the instruction set with the control interfaces of each system, and filter conflicting instructions; Instructions are distributed in a time sequence to gas-fired boilers, primary condensing waste heat recovery systems, secondary whitening heat exchange systems, four-pipe air source heat pump units, underfloor heating systems, and air conditioning heating systems.

6. The method for optimizing energy scheduling of natural gas boiler flue gas whitening and waste heat recovery according to claim 5, characterized in that, Verifying the compatibility of the instruction set with the control interfaces of each system includes: Extract the communication protocols and data format specifications of each system; Compare each control parameter in the instruction set with the allowable range specified in the standard. Mark the out-of-limit instruction and replace it with a preset safe value.

7. The method for optimizing energy scheduling of natural gas boiler flue gas whitening and waste heat recovery according to claim 6, characterized in that, Time-series dispatch instructions include: Based on the response latency characteristics and instruction execution duration of each system, allocate instruction sending timing; A timing synchronization mechanism is employed to ensure the coordination of instruction execution across multiple systems; Record instruction distribution logs and update them to the energy dispatch database in real time.

Images