Whole house cooling and heating intelligent decision-making system and method based on dynamic environment perception

Abstract

The invention relates to the field of heating ventilation and air conditioning engineering, discloses a whole-house cooling and heating intelligent decision-making system and method based on dynamic environment perception, and aims to solve the problem of low efficiency of traditional fixed threshold value control. The method comprises the following steps: collecting whole house environment parameters and equipment state data through a sensor network; identifying a dynamic perturbation event based on the sliding time window; updating the building thermal characteristic model; constructing a multi-objective optimization decision framework to generate an optimal control strategy; and verifying the strategy performance in the high-fidelity simulation model and performing adaptive adjustment. The system comprises an environment sensing module, an equipment state monitoring module, a dynamic disturbance identification module, a building thermal characteristic modeling module, a multi-objective optimization decision module, a strategy simulation verification module and an online learning and adaptive adjustment module. By the adoption of the technical scheme, accurate energy regulation and control can be achieved, the comprehensive energy efficiency of the system is improved by 15%-25%, and the thermal comfort reaching rate is improved to 95% or above.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of heating, ventilation and air conditioning engineering, and particularly relates to a whole-house cooling and heating intelligent decision system and method based on dynamic environment perception. BACKGROUND

[0002] With the continuous evolution of smart home and building energy management technology, the intelligent level of whole-house cooling and heating system has become an important indicator to measure the living comfort and energy utilization efficiency. Various environmental regulation devices are widely deployed in modern residences, including air conditioners, floor heating, fresh air systems, and intelligent temperature controllers, etc., and their collaborative operation directly affects the indoor thermal comfort and overall energy consumption performance. However, the current mainstream cooling and heating control strategy still highly depends on preset temperature thresholds or simple time scheduling logic, lacks comprehensive perception and response ability to multi-dimensional dynamic factors such as building thermal characteristics, external meteorological conditions and user behavior, and is difficult to achieve fine and adaptive energy regulation in complex and variable living environment.

[0003] Among them, the whole-house cooling and heating intelligent decision technology based on dynamic environment perception aims to break through the limitations of traditional control mode, and by fusing building physical model, real-time environment data and device operation state, an intelligent regulation system with prediction and optimization ability is constructed. The core of this technology direction is how to efficiently model the complex coupling relationship between devices, and quickly generate a multi-objective control strategy that takes into account comfort, energy saving and device life under dynamic disturbance, so as to realize the paradigm shift from "passive response" to "active decision".

[0004] The existing technology still faces multiple challenges in achieving the above objectives: first, most systems only use static rules or isolated device control logic, which cannot capture the spatio-temporal evolution law of the whole-house thermal environment; second, there is a lack of real-time perception and feedforward compensation mechanism for meteorological mutations, door and window opening and closing, personnel flow and other dynamic disturbances, resulting in regulation lag or over-regulation; third, the existing strategy generation methods are mostly based on simplified assumptions or local optimization, and it is difficult to balance multi-device collaboration, multi-objective trade-off and long-term operation stability; finally, the strategy evaluation link generally lacks high-fidelity simulation and online learning ability, and cannot effectively verify the strategy performance or continuously optimize the decision model before real deployment. These problems are particularly prominent in the context of frequent extreme weather and increasing user demand for individualization, seriously restricting the energy efficiency improvement and user experience optimization of smart home energy systems, and there is an urgent need for a new whole-house cooling and heating decision paradigm that integrates dynamic perception, intelligent modeling and closed-loop optimization. SUMMARY

[0005] The purpose of the present application is to make up for the shortcomings of the prior art, and to provide a whole-house cooling and heating intelligent decision system and method based on dynamic environment perception, which can effectively solve the problems in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: In one aspect, a whole-house intelligent decision-making system for cooling and heating based on dynamic environmental perception, the system comprising the following components:

[0007] The environmental sensing module is used to collect multi-dimensional environmental data of the whole house in real time. The multi-dimensional environmental data includes indoor temperature distribution, humidity distribution, air velocity, door and window opening and closing status, personnel distribution density, and outdoor meteorological parameters.

[0008] The equipment status monitoring module is used to continuously acquire the operating status parameters of the cooling and heating equipment. The operating status parameters include compressor frequency, water pump speed, valve opening, instantaneous energy consumption value and cumulative value.

[0009] The dynamic disturbance identification module connects the environmental perception module and the equipment status monitoring module. It is used to identify sudden dynamic disturbance events based on the environmental data sequence and equipment status sequence within a sliding time window.

[0010] The building thermal characteristics modeling module includes built-in building envelope thermal resistance parameters, spatial geometric topology relationships, and material heat capacity coefficients to construct a whole-house thermal dynamic response model.

[0011] A multi-objective optimization decision-making module is coupled with a building thermal characteristic modeling module and a dynamic disturbance identification module to generate a coordinated equipment control strategy in the future regulation time domain based on a model predictive control framework.

[0012] The strategy simulation verification module receives the control strategy output by the multi-objective optimization decision module, performs forward simulation in a high-fidelity building thermal model, and predicts the thermal comfort index and system energy consumption after the strategy is executed.

[0013] The online learning and adaptive adjustment module connects the strategy simulation verification module with the actual equipment execution unit. By comparing the deviation between predicted performance and actual performance, it dynamically corrects the building thermal characteristic model parameters and control strategy weight coefficients.

[0014] Preferably, the environmental sensing module is deployed with a distributed wireless sensor network, where sensor nodes communicate via the 2.4GHz ZigBee protocol with a sampling frequency of not less than 0.1Hz; the indoor temperature sensor uses a DS18B20 digital temperature sensor with a measurement accuracy of ±0.5℃; the humidity sensor uses an SHT35 integrated sensor with a measurement range of 0-100%RH and an accuracy of ±2%RH; the door and window status sensor uses a Hall effect magnetic switch; and the personnel distribution detection uses passive positioning technology based on a pyroelectric infrared sensor array with a spatial resolution of 1 square meter.

[0015] Furthermore, the dynamic disturbance identification module employs a mutation detection algorithm based on sliding window standard deviation analysis, defining a time window length T = 300 seconds and a sliding step size of 10 seconds; the formula for calculating the standard deviation σ of the temperature sequence is: Where N is the number of sampling points within the window, and T i Let be the temperature of the i-th sampling point. The average temperature within the window; the rate of temperature change. The calculation is performed using the first-order backward difference method: Wherein is the sampling interval (60 seconds). When 3 consecutive sampling points... When the temperature rises by ℃ / min and σ > 1.2℃, it is considered a dynamic disturbance event.

[0016] For disturbances caused by the opening and closing of doors and windows, state duration filtering is used, and the opening and closing state lasting less than 30 seconds is considered an invalid disturbance.

[0017] Furthermore, the building thermal characteristic modeling module uses the lumped parameter method to establish the room heat balance equation, the core differential equation of which is expressed as:

[0018]

[0019] Where C is the indoor air heat capacity, and T i For indoor temperature, U ij T is the heat transfer coefficient between room i and adjacent area j. j Q represents the temperature of the adjacent region. internal Q represents the heat generated by the indoor heat source. HVAC This refers to the heating or cooling capacity supplied to the HVAC system. Q is the unit of measurement for this system. internal Estimated using indoor occupant density sensor data: Q internal = 0.08 × P × A (P is the number of people, A is the area of ​​the region, unit: W / m²) 2 );Q HVAC According to the compressor power P of the equipment status monitoring module comp and water pump power P pump Calculate: Q HVAC =COP×(P comp +P pump COP is the equipment energy efficiency ratio (preset value 2.8).

[0020] Preferably, the multi-objective optimization decision module constructs a multi-objective cost function that includes thermal comfort deviation, total system energy consumption, and equipment lifespan loss, and its mathematical expression is:

[0021]

[0022] Where α, β, and γ are weighting coefficients, and Tactual T represents the actual temperature. setpoint To set the temperature, P total Let Dx be the total power of the system and Dx be the operating loss factor of the k-th device. The optimization process uses a constrained non-dominated sorting genetic algorithm to solve for the Pareto optimal solution set.

[0023] Furthermore, the strategy simulation verification module integrates a high-fidelity building energy consumption simulation engine based on the EnergyPlus kernel, with a simulation step size of 1 minute, which can accurately simulate the effects of wall heat storage, solar radiation heat gain, and air flow on the indoor thermal environment; the simulation output includes the predicted average voting index PMV, local dissatisfaction rate PPD, and sub-item energy consumption data.

[0024] In addition, the online learning and adaptive adjustment module uses the recursive least squares method to identify the building thermal model parameters online and performs a model parameter update every 24 hours; the control strategy weight coefficients are dynamically adjusted based on the multi-objective performance evaluation of historical operating data, with an adjustment cycle of 7 days.

[0025] On the other hand, a whole-house intelligent decision-making method for cooling and heating based on dynamic environmental perception, the specific steps of which are as follows:

[0026] Step S110: Periodically collect whole-house environmental parameters and equipment operating status data through a sensor network deployed inside and outside the building;

[0027] Step S120: Analyze the temporal characteristics of environmental data based on a sliding time window, and identify three typical dynamic disturbance events in real time: opening and closing of doors and windows, movement of people, and sudden changes in weather.

[0028] Step S130: Call the pre-stored physical parameters of the building envelope and dynamically update the state variables of the building thermal characteristic model in combination with the real-time collected environmental data.

[0029] Step S140: Construct a multi-objective cost function with thermal comfort, system energy consumption and equipment life as optimization objectives, and use a model predictive control framework to solve for the optimal equipment control sequence in the control time domain for the next 2 hours;

[0030] Step S150: Input the generated control strategy into the high-fidelity building thermal simulation model to predict the evolution of the indoor thermal environment and the distribution of system energy consumption after the strategy is implemented.

[0031] Step S160: Compare the compliance of the strategy simulation results with the preset performance indicators. Issue the verified strategies to the device execution unit, and return to step S140 to re-optimize the strategies that do not meet the standards.

[0032] Step S170: Based on the deviation between actual operating data and simulation prediction data, a parameter identification algorithm is used to correct the building thermal model, and the multi-objective weight coefficients are adaptively adjusted according to long-term performance statistics.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] By integrating a distributed environmental sensing network with a dynamic disturbance identification mechanism, the system accurately captures the spatiotemporal evolution of the whole-house thermal environment and sudden disturbances, solving the problem of lagging regulation caused by the single sensing dimension in traditional systems.

[0035] By adopting a technical approach that combines physical mechanism-based modeling of building thermal characteristics with data-driven online learning, the adaptability of the system and the accuracy of model prediction under different building structures and climatic conditions have been significantly improved.

[0036] By introducing a multi-objective optimization decision-making framework and a high-fidelity strategy simulation verification process, the overall performance can be quantitatively evaluated before the strategy is deployed, effectively avoiding energy waste or decreased comfort caused by strategy defects.

[0037] A complete technological closed loop has been constructed, encompassing perception, decision-making, verification, and self-learning. This has enabled a paradigm shift in cooling and heating systems from static rule-based control to dynamic intelligent decision-making, resulting in a 15%-25% improvement in overall system energy efficiency and an increase in thermal comfort compliance rate to over 95%. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of the overall technical solution architecture of the whole-house intelligent decision-making system for cooling and heating based on dynamic environmental perception proposed in this invention.

[0039] Figure 2 This is a schematic diagram of the core principle framework of dynamic disturbance identification and multi-objective optimization decision-making in this invention;

[0040] Figure 3 This is a logical flowchart of the environmental perception and building thermal characteristic modeling in this invention.

[0041] Figure 4 This is a schematic diagram of the multi-level interaction relationship and data flow between strategy simulation verification and online learning adaptive adjustment in this invention;

[0042] Figure 5 This is a flowchart illustrating the overall process framework of the whole-house intelligent decision-making method for cooling and heating in this invention. Detailed Implementation

[0043] To further illustrate the technical means and effects adopted by the present invention to achieve the intended purpose, the specific embodiments according to the present invention will be described in detail below with reference to the accompanying drawings and preferred embodiments.

[0044] Example 1

[0045] In a 180-square-meter detached house in northern China, this system was deployed to achieve fully automated intelligent control of heating in winter and cooling in summer. The house has a north-south facing structure and includes multiple functional areas such as a living room, bedrooms, study, and kitchen. Its exterior walls are made of 200mm thick aerated concrete blocks, and the windows are double-glazed PVC windows. The system captures multi-dimensional environmental parameters throughout the house in real time through an environmental sensing module. This module is deployed with a distributed wireless sensor network using the 2.4GHz ZigBee protocol, with a total of 15 sensor nodes evenly distributed throughout the rooms and corridors, and a sampling frequency set to 0.1 Hz. Indoor temperature measurement uses a DS18B20 digital temperature sensor with an accuracy of ±0.5 degrees Celsius, uploading data to a centralized processor every 30 seconds. Humidity monitoring uses an SHT35 integrated sensor, covering a range of 0 to 100% relative humidity with an accuracy of ±2% relative humidity, and data is updated every 60 seconds. Door and window status sensing utilizes Hall effect magnetic switches, installed on all exterior windows and entrance door frames, capable of real-time detection of opening and closing status and recording the duration of each status. Personnel distribution detection employs passive positioning technology based on a pyroelectric infrared sensor array. Sensors are installed in a grid pattern on the ceiling, with each sensor covering an area of ​​1 square meter, outputting estimated personnel presence and density values ​​for each area at 1-minute intervals. Outdoor meteorological parameters are obtained through an internet-connected meteorological data service interface, including outdoor dry-bulb temperature, relative humidity, solar radiation intensity, and wind speed and direction, updated every 5 minutes.

[0046] The equipment status monitoring module interacts with the residential variable frequency multi-split air conditioning system, floor radiant heating system, and fresh air unit via the Modbus-RTU communication protocol. This module continuously collects the compressor operating frequency with an accuracy of 0.1 Hz and a collection cycle of 10 seconds; monitors the water pump speed, measuring from 0 to 3000 rpm with an accuracy of ±5 rpm; reads the opening signals of the electric two-way and three-way valves with an opening resolution accuracy of 1%; and simultaneously calculates the system's instantaneous power consumption and cumulative energy consumption in real time. Power measurement uses a Hall current sensor with an accuracy class of 1.0, and data is recorded every 15 seconds. All equipment status parameters are timestamped and stored in a circular buffer with a capacity of 24 hours of operating data.

[0047] The dynamic disturbance identification module receives data streams from the environmental perception module and the equipment status monitoring module, and uses a mutation detection algorithm based on sliding window standard deviation analysis for real-time disturbance identification. This module defines a time window length T of 300 seconds and calculates the standard deviation σ of the temperature sequence within the window at each sampling moment. When the system detects that the temperature change rate at three consecutive sampling points exceeds the threshold of 0.5 degrees Celsius per minute and the instantaneous standard deviation σ is greater than 1.2 degrees Celsius, it is automatically identified as a temperature dynamic disturbance event. For door and window opening and closing disturbances, the module uses a state duration filtering algorithm; any change in opening or closing state duration less than 30 seconds is considered an invalid disturbance and does not trigger a system response. In specific implementation, when the living room sliding door is open for more than 30 seconds, the module immediately identifies the disturbance event and extracts the event type as door / window opening / closing, the location as the living room, the duration as from the moment the door opens to the moment it closes, and the affected area as the living room and adjacent corridor area.

[0048] The building thermal performance modeling module incorporates a database of thermal resistance parameters for the building envelope of this residence, including the heat transfer coefficients of 0.45 W / m² Kelvin for exterior walls, 2.2 W / m² Kelvin for exterior windows, and 0.35 W / m² Kelvin for the roof. The module also stores the spatial geometric topology of each room, including room volume, shared wall area between adjacent rooms, door and window dimensions, and azimuth. The material heat capacity database records the specific heat capacity of concrete as 0.92 kJ / kg Kelvin, wood as 1.26 kJ / kg Kelvin, and air as 1.005 kJ / kg Kelvin. Based on these parameters, the module uses the lumped parameter method to establish a whole-house thermal dynamic response model, the core of which is expressed as:

[0049]

[0050] Where C is the indoor air heat capacity, and T i For indoor temperature, U ij T is the heat transfer coefficient between room i and adjacent area j. j Q represents the temperature of the adjacent region. internal Q represents the heat generated by the indoor heat source. HVAC This model provides heating or cooling capacity for the HVAC system. It performs numerical solutions in 5-minute increments, updating the predicted temperature values ​​for each room in real time.

[0051] The multi-objective optimization decision-making module couples the building thermal characteristic modeling module and the dynamic disturbance identification module, generating a coordinated equipment control strategy for the next two hours of regulation based on a model predictive control framework. This module constructs a multi-objective cost function that includes thermal comfort deviation, total system energy consumption, and equipment lifespan loss, mathematically expressed as:

[0052]

[0053] Where α, β, and γ are weighting coefficients, and T actual T represents the actual temperature. setpoint To set the temperature, P total For the total power of the system, D k Let be the operating loss factor of the kth device.

[0054] The optimization process uses a constrained non-dominated sorting genetic algorithm to solve for the Pareto optimal solution set: 1) Initialize the population size to 100; 2) Define the fitness function as the reciprocal of the cost function J; 3) Non-dominated sorting is based on stratification according to the objective function value, and sorting within the same stratum according to crowding; 4) Constraint handling uses a penalty function method: for violations of the upper limit of device frequency (f max For individuals with a frequency of 50Hz, the cost function is increased by a penalty term λ×(ff). max ) 2 (λ=10); 5) The crossover operation uses simulated binary crossover with a probability of 0.8; the mutation operation uses Gaussian mutation with a probability of 0.1. After 200 iterations, the Pareto front optimal solution is output.

[0055] The optimization variables include compressor frequency setpoint, water pump speed setpoint, valve opening setpoint, and fresh air unit start / stop status. The constraints include upper and lower limits of equipment operating parameters, temperature comfort range, and power limits.

[0056] The strategy simulation verification module receives the control strategy output by the multi-objective optimization decision module and inputs it into a high-fidelity building energy consumption simulation engine based on the EnergyPlus kernel for forward simulation. The simulation step size is set to 1 minute, and the simulation duration covers the entire control time domain of 2 hours. The simulation engine accurately simulates the wall heat storage effect, considering material heat capacity and heat transfer delay; calculates solar radiation heat gain based on building orientation, window area, and shading coefficient; and simulates the impact of airflow on the indoor thermal environment using a multi-region airflow network model. The simulation output includes the predicted average voting index (PMV) with a calculation accuracy of 0.01; the local dissatisfaction rate (PPD) with a calculation accuracy of 0.1%; and system component energy consumption data, including compressor power consumption, water pump power consumption, and fan power consumption, with an accuracy of 1 watt-hour.

[0057] The online learning and adaptive adjustment module connects the strategy simulation verification module with the actual equipment execution unit, dynamically correcting system parameters by comparing the deviation between predicted and actual performance. This module uses the recursive least squares method to identify building thermal model parameters online, updating the model parameters every 24 hours for 30 minutes, during which time the system maintains the current control strategy. The control strategy weight coefficients are dynamically adjusted based on multi-objective performance evaluation using historical operating data, with an adjustment cycle of 7 days. Evaluation indicators include average temperature deviation, cumulative energy consumption, and equipment operating time distribution. The adjustment range of the weight coefficients is limited to ±20% to ensure the stability of system control.

[0058] The specific process of the system's whole-house intelligent decision-making method for cooling and heating is as follows: Step S110: Periodically collect whole-house environmental parameters and equipment operating status data through a sensor network deployed inside and outside the building. The data collection cycle is 30 seconds, including 15 temperature measurement points, 12 humidity measurement points, 8 door and window status monitoring points, 6 personnel distribution areas, and 4 types of equipment operating parameters. Step S120: Analyze the temporal characteristics of environmental data based on a sliding time window. The time window length is 5 minutes, and the sliding step is 30 seconds. Real-time identification of three typical dynamic disturbance events: door and window opening and closing, personnel flow, and sudden weather changes, with an identification response time of less than 10 seconds. Step S130: Call the pre-stored physical parameters of the building envelope and combine them with the real-time collected environmental data to dynamically update the state variables of the building thermal characteristic model. The update frequency is 5 minutes, and the state variables include the temperature of each room, the surface temperature of the walls, and the air humidity. Step S140 constructs a multi-objective cost function with thermal comfort, system energy consumption, and equipment lifespan as optimization objectives. A model predictive control framework is used to solve for the optimal equipment control sequence within the next 2 hours of regulation. The optimization calculation time is 3 minutes, outputting equipment setpoints for 24 control periods. Step S150 inputs the generated control strategy into a high-fidelity building thermal simulation model to predict the evolution of the indoor thermal environment and the distribution of system energy consumption after the strategy is implemented. The simulation duration is 2 hours, outputting state prediction values ​​at 1440 time points. Step S160 compares the consistency of the strategy simulation results with preset performance indicators, including a PMV value between -0.5 and +0.5, a PPD value below 10%, and energy consumption not exceeding 110% of the baseline value. Validated strategies are immediately implemented to the equipment execution units; strategies that do not meet the standards return to step S140 for re-optimization, with a maximum of 3 repetitions. Step S170 uses a parameter identification algorithm to correct the building thermal model based on the deviation between actual operating data and simulation prediction data. The parameter identification cycle is 24 hours. The multi-objective weight coefficients are adaptively adjusted based on long-term performance statistics. The weight adjustment cycle is 7 days.

[0059] Example 2

[0060] In a single residential unit of a high-rise apartment building in southern China, this system was configured for precise cooling control under hot and humid summer conditions. The apartment unit has a floor area of ​​85 square meters, faces east and west, and its main heat load comes from afternoon sun exposure and indoor occupancy. The system's environmental sensing module was optimized for this application scenario. Two temperature sensors were added to the west-facing room, increasing the sampling frequency to 0.2 Hz; three pyroelectric infrared sensors were added to the frequently used living room and bedroom areas, improving the spatial resolution to 0.5 square meters; and two additional carbon dioxide concentration sensors were deployed to monitor indoor air quality. The equipment status monitoring module connects to a variable frequency split air conditioner and a dehumidifier, monitoring parameters including compressor frequency, indoor and outdoor fan speeds, electronic expansion valve opening, dehumidifier compressor start / stop status, and condensate discharge status.

[0061] The dynamic disturbance identification module enhances its ability to identify meteorological abrupt changes, specifically tailored to the climate characteristics of southern China. It immediately triggers a meteorological abrupt disturbance event when the outdoor temperature rises by more than 3 degrees Celsius within one hour or the relative humidity increases by more than 15% within 30 minutes. The building thermal characteristic modeling module, designed for apartment structures, focuses on optimizing the solar radiation heat gain model on the west wall, considering the impact of solar altitude and azimuth angles on radiation intensity at different times. The model accuracy is 12% higher than in Example 1. The multi-objective optimization decision module adds a dehumidification energy consumption term to the cost function, with a weighting coefficient γ set to 0.15. The optimization variables are expanded to include the air conditioning temperature setpoint, fan speed, and dehumidifier start-stop cycle. The strategy simulation verification module employs a more refined humidity transfer model, capable of predicting changes in indoor dew point temperature and condensation risk. The online learning and adaptive adjustment module shortens the parameter update cycle to 12 hours to adapt to the frequent changes in meteorological conditions in southern China.

[0062] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A whole-house intelligent decision-making system for cooling and heating based on dynamic environmental perception, characterized in that, The system includes the following components: The environmental sensing module is used to collect multi-dimensional environmental data of the whole house in real time. The multi-dimensional environmental data includes indoor temperature distribution, humidity distribution, air velocity, door and window opening and closing status, personnel distribution density, and outdoor meteorological parameters. The equipment status monitoring module is used to continuously acquire the operating status parameters of the cooling and heating equipment. The operating status parameters include compressor frequency, water pump speed, valve opening, instantaneous energy consumption value and cumulative value. The dynamic disturbance identification module connects the environmental perception module and the equipment status monitoring module. It is used to identify sudden dynamic disturbance events based on the environmental data sequence and equipment status sequence within a sliding time window. The building thermal characteristics modeling module includes built-in building envelope thermal resistance parameters, spatial geometric topology relationships, and material heat capacity coefficients to construct a whole-house thermal dynamic response model. A multi-objective optimization decision-making module is coupled with a building thermal characteristic modeling module and a dynamic disturbance identification module to generate a coordinated equipment control strategy in the future regulation time domain based on a model predictive control framework. The strategy simulation verification module receives the control strategy output by the multi-objective optimization decision module, performs forward simulation in a high-fidelity building thermal model, and predicts the thermal comfort index and system energy consumption after the strategy is executed. The online learning and adaptive adjustment module connects the strategy simulation verification module with the actual equipment execution unit. By comparing the deviation between predicted performance and actual performance, it dynamically corrects the building thermal characteristic model parameters and control strategy weight coefficients.

2. The whole-house intelligent decision-making system for cooling and heating based on dynamic environmental perception as described in claim 1, characterized in that, The environmental sensing module is equipped with a distributed wireless sensor network. The sensor nodes communicate via the 2.4GHz ZigBee protocol, with a sampling frequency of no less than 0.1Hz. The indoor temperature sensor uses a DS18B20 digital temperature sensor with a measurement accuracy of ±0.5℃. The humidity sensor uses an SHT35 integrated sensor with a measurement range of 0-100%RH and an accuracy of ±2%RH. The door and window status sensor uses a Hall effect magnetic switch. The personnel distribution detection uses passive positioning technology based on a pyroelectric infrared sensor array, with a spatial resolution of 1 square meter.

3. The whole-house intelligent decision-making system for cooling and heating based on dynamic environmental perception as described in claim 1, characterized in that, The dynamic disturbance identification module employs a mutation detection algorithm based on sliding window standard deviation analysis, defining a time window length T = 300 seconds and a sliding step size of 10 seconds; the formula for calculating the temperature series standard deviation σ is: Where N is the number of sampling points within the window, and T i Let be the temperature of the i-th sampling point. The average temperature within the window; the rate of temperature change. The calculation is performed using the first-order backward difference method: in The sampling interval is 60 seconds. When three consecutive sampling points... Furthermore, when σ > 1.2℃, it is determined to be a dynamic disturbance event. For door and window opening and closing disturbances, state duration filtering is used, and the opening and closing state lasting less than 30 seconds is considered an invalid disturbance.

4. The whole-house intelligent decision-making system for cooling and heating based on dynamic environmental perception as described in claim 1, characterized in that, The building thermal characteristics modeling module uses the lumped parameter method to establish the room heat balance equation, and its core differential equation is expressed as follows: Where C is the indoor air heat capacity, and T i For indoor temperature, U ij T is the heat transfer coefficient between room i and adjacent area j. j Q represents the temperature of the adjacent region. internal Q represents the heat generated by the indoor heat source. HVAC Provides heating or cooling capacity for HVAC systems. Q internal Estimated using indoor occupant density sensor data: Q internal = 0.08 × P × A (P is the number of people, A is the area of ​​the region, unit: W / m²) 2 );Q HVAC According to the compressor power P of the equipment status monitoring module comp and water pump power P pump Calculate: Q HVAC =COP×(P comp +P pump COP is the equipment energy efficiency ratio (preset value 2.8).

5. The whole-house intelligent decision-making system for cooling and heating based on dynamic environmental perception according to claim 1, characterized in that, The multi-objective optimization decision module constructs a multi-objective cost function that includes thermal comfort deviation, total system energy consumption, and equipment lifespan loss. Its mathematical expression is as follows: Where α, β, and γ are weighting coefficients, and T actual T represents the actual temperature. setpoint To set the temperature, P total For the total power of the system, D k Let be the operating loss factor of the k-th device; the optimization process uses a constrained non-dominated sorting genetic algorithm to solve for the Pareto optimal solution set.

6. The whole-house intelligent decision-making system for cooling and heating based on dynamic environmental perception according to claim 1, characterized in that, The strategy simulation verification module integrates a high-fidelity building energy consumption simulation engine based on the EnergyPlus kernel. The simulation step size is 1 minute, which can accurately simulate the effects of wall heat storage, solar radiation heat gain and air flow on the indoor thermal environment. The simulation output includes the predicted average voting index PMV, local dissatisfaction rate PPD and sub-item energy consumption data.

7. The whole-house intelligent decision-making system for cooling and heating based on dynamic environmental perception according to claim 1, characterized in that, The online learning and adaptive adjustment module uses the recursive least squares method to identify the building thermal model parameters online and performs a model parameter update every 24 hours; the control strategy weight coefficients are dynamically adjusted based on the multi-objective performance evaluation of historical operating data, with an adjustment cycle of 7 days.

8. A whole-house intelligent decision-making method for cooling and heating based on dynamic environmental perception, characterized in that, The method includes the following steps: The S110 periodically collects environmental parameters and equipment operating status data throughout the building through a sensor network deployed inside and outside the building. S120 analyzes the temporal characteristics of environmental data based on a sliding time window, and identifies three typical dynamic disturbance events in real time: opening and closing of doors and windows, movement of people and sudden weather changes. S130: Call the pre-stored physical parameters of the building envelope and combine them with the real-time collected environmental data to dynamically update the state variables of the building thermal characteristic model. S140, construct a multi-objective cost function with thermal comfort, system energy consumption and equipment life as optimization objectives, and use a model predictive control framework to solve the optimal equipment control sequence in the control time domain for the next 2 hours; S150 inputs the generated control strategy into the high-fidelity building thermal simulation model to predict the evolution of the indoor thermal environment and the distribution of system energy consumption after the strategy is executed. S160: Compare the consistency between the simulation results of the strategy and the preset performance indicators. Issue the verified strategies to the equipment execution unit, and return to step S140 to re-optimize the strategies that do not meet the standards. S170 uses a parameter identification algorithm to correct the building thermal model based on the deviation between actual operating data and simulation prediction data, and adaptively adjusts the multi-objective weight coefficients based on long-term performance statistics.

9. The intelligent decision-making method for whole-house cooling and heating based on dynamic environmental perception according to claim 8, characterized in that, In step S120, the time window length is 5 minutes, the sliding step size is 30 seconds, and the identification response time is less than 10 seconds; in step S140, the optimization calculation time is 3 minutes, and the equipment setting values ​​for 24 control time periods are output; in step S150, the simulation duration is 2 hours, and the state prediction values ​​for 1440 time points are output; in step S160, the performance indicators include a PMV value between -0.5 and +0.5, a PPD value below 10%, and energy consumption not exceeding 110% of the baseline value, and the optimization is repeated a maximum of 3 times.

10. The intelligent decision-making method for whole-house cooling and heating based on dynamic environment perception according to claim 8, characterized in that, In step S170, the parameter identification period is 24 hours, the weight adjustment period is 7 days, and the weight coefficient adjustment range is limited to ±20%.

Images