Intelligent energy-saving building ventilation system self-adaptive to environment change

Abstract

The invention provides an intelligent energy-saving building ventilation system self-adaptive to environmental changes, and relates to the technical field of intelligent buildings. The system comprises a hardware layer and a functional layer, the hardware layer comprises a sensor group, an actuator and a controller, and the functional layer is composed of a plurality of program modules operated by the controller. Wherein the data sensing and processing module is responsible for collecting indoor and outdoor environment parameters and preprocessing the data; the deep learning prediction module predicts a future environmental parameter trend and system energy consumption by using historical environmental data, an equipment operation state and external environmental data on the basis of an LSTM (Long Short Term Memory) network; the fuzzy logic control module combines the real-time environment data and the prediction data, and generates a preliminary control instruction according to a fuzzy rule. According to the system, through cooperative work of all the modules, intelligent adjustment and energy-saving optimization of the building ventilation system are achieved, energy consumption is reduced, and meanwhile the indoor comfort degree and the air quality are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent buildings, in particular to an intelligent energy-saving building ventilation system that is self-adaptive to environmental changes. BACKGROUND

[0002] A building ventilation system is a key link for maintaining indoor air quality and ensuring personnel health and comfort, especially in large public buildings and office spaces, and its energy consumption accounts for a large proportion of the total building energy consumption. With the deepening of the concepts of intelligent buildings and green buildings, the requirements for the ventilation system are no longer limited to basic ventilation functions, but need to achieve fine, self-adaptive and energy-saving operation on the premise of meeting indoor environmental quality. Traditional ventilation systems mostly use constant air volume or control methods based on simple threshold values (such as CO2 concentration), which are difficult to cope with complex working conditions such as dynamic changes in the number of building users, fluctuations in external environment, and multi-objective collaborative optimization.

[0003] In the prior art, some attempts have been made to improve the intelligence and energy saving of the ventilation system. For example, the intelligent energy-saving building ventilation system described in patent CN115406073A obtains corridor pictures and gas data through a data acquisition end, analyzes them through a processing center, and controls the ventilation volume according to the flow of people and gas parameters. This system achieves on-demand ventilation to some extent, but still has significant limitations: first, its control is based on historical mean data processing, with a lagging response, and cannot make forward-looking adjustments to environmental changes; second, it relies on fixed formulas and empirical thresholds, and lacks real-time learning and optimization capabilities for nonlinear and time-varying building ventilation systems; third, it does not introduce external environmental parameters (such as weather and electricity prices), making it difficult to achieve true energy-saving optimization under complex and changing conditions; fourth, the control strategy is relatively rigid, cannot integrate user subjective comfort feedback, and does not have a mechanism for strategy verification and iteration through digital twinning and other technologies. SUMMARY

[0004] (I) Technical problems solved In view of the deficiencies of the prior art, the present application provides an intelligent energy-saving building ventilation system that is self-adaptive to environmental changes, which solves the problems raised in the background art.

[0005] (II) Technical solutions To achieve the above purpose, the present application is implemented by the following technical solutions: an intelligent energy-saving building ventilation system that is self-adaptive to environmental changes, comprising: a hardware layer and a functional layer, the hardware layer comprising a sensor group, an actuator, and a controller in communication with the sensor group and the actuator; the functional layer is realized by a program module run by the controller, comprising: a data perception and processing module configured to obtain indoor and outdoor environmental parameters from the sensor group and to preprocess the data including denoising, filtering and normalization; a deep learning prediction module configured to receive pre-processed data from the data perception and processing module, and to establish a time series prediction model based on an LSTM network, with historical environmental data, equipment operating status and external environmental data as inputs, and with environmental parameter trends and system energy consumption prediction values in a future period as outputs; a fuzzy logic control module configured to receive real-time environmental data from the data perception and processing module, and to receive prediction data from the deep learning prediction module, to convert the data into fuzzy language variables by comparing the data with pre-set quantization ranges, and to generate preliminary control instructions according to a fuzzy rule base; a global optimization module configured to receive prediction data from the deep learning prediction module, to solve a multi-objective optimization problem by an optimization algorithm, to generate a globally optimal operating strategy, and to dynamically adjust the set values or rule parameters of the fuzzy logic control module based on the strategy; The controller is configured to send the control instructions generated by the fuzzy logic control module to the actuators to drive the ventilation equipment to operate.

[0006] Preferably, the sensor group includes an indoor CO2 concentration sensor, a temperature and humidity sensor, a personnel counter, and an outdoor temperature and humidity sensor. The system is also provided with an external data interface for real-time acquisition of weather forecasts, electricity price information, and air quality index data with a resolution of no less than 1 hour.

[0007] Preferably, the actuators include a variable frequency fan, a damper actuator, and a table cooler valve controller. The control instructions output by the controller include speed set values of the variable frequency fan at 20%–100% of the rated speed, opening instructions of the regional dampers, and opening instructions of the table cooler valves.

[0008] Preferably, the data perception and processing module is configured to perform denoising, filtering, and normalization preprocessing on the indoor and outdoor environmental parameters collected by the sensor group, and to provide standardized data streams to the deep learning prediction module and the fuzzy logic control module.

[0009] Preferably, the deep learning prediction module constructs a time series prediction model based on an LSTM network. The time window length N of the historical environmental data input to the model is 12–168, and the model outputs environmental parameter trends and system energy consumption prediction values in a future period of M=1–24.

[0010] Preferably, the fuzzy logic control module uses an adaptive neuro-fuzzy inference system (ANFIS) to update membership functions and fuzzy rule parameters online according to historical operating data, and to convert real-time environmental data and prediction data into fuzzy language variables to generate preliminary control instructions.

[0011] Preferably, the global optimization module supports both genetic algorithm and reinforcement learning modes: In the genetic algorithm mode, the population size is 50-200, the crossover probability is 0.7-0.9, and the mutation probability is 0.01-0.1; in the reinforcement learning mode, the agent reward function comprehensively considers the energy consumption cost, thermal comfort, and indoor air quality, and the weight of each index can be adjusted in real time according to the external power grid demand response signal or the price increase.

[0012] Preferably, the system further integrates a digital twin module for constructing a high-fidelity virtual model of the ventilation system and the building; the operation strategy generated by the global optimization module is first simulated and verified in the digital twin module, and then iteratively optimized before being executed in the physical system.

[0013] Preferably, the human-computer interaction module receives user subjective comfort feedback, and the global optimization module and the fuzzy logic control module dynamically adjust comfort-related parameters and set values based on the feedback data to continuously improve user experience and system robustness.

[0014] (Three) beneficial effects The present application provides an intelligent energy-saving building ventilation system that can adapt to environmental changes, which has the following beneficial effects: Through the collaborative work of the deep learning prediction module and the global optimization module, the system can accurately predict and dynamically optimize environmental parameters and equipment energy consumption. The system can intelligently adjust the operation strategy according to the indoor and outdoor environmental changes in different time periods, reduce unnecessary equipment running time and power output, and thus significantly reduce energy consumption.

[0015] The system uses the fuzzy logic control module to dynamically adjust the indoor temperature, humidity, and carbon dioxide concentration based on real-time environmental data provided by sensors and future trend data from the deep learning prediction module. Through adaptive algorithms, the system controls indoor temperature fluctuations within ±0.5℃ of the set value, and the proportion of time when carbon dioxide concentration exceeds the standard is reduced to less than 1%. At the same time, the system also supports receiving user subjective comfort feedback, further optimizing the operation strategy, and significantly improving user experience, solving the problem of slow response to environmental changes, insufficient comfort and air quality protection of traditional systems. BRIEF DESCRIPTION OF DRAWINGS

[0016] Fig. 1 The figure is a schematic diagram of the system architecture of the present application; Fig. 2 The figure is a schematic diagram of the system flow of the present application. DETAILED DESCRIPTION

[0017] 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. Example

[0018] like Figs. 1-2 As shown, this embodiment of the invention provides an intelligent energy-saving building ventilation system that adapts to environmental changes. The system consists of a hardware layer and a functional layer. The hardware layer includes a sensor group, actuators, and a controller communicatively connected to the sensor group and actuators. The sensor group includes indoor temperature and humidity sensors, carbon dioxide concentration sensors, and occupancy counters deployed in various areas of the building, as well as outdoor temperature and humidity sensors installed on the building's exterior walls. The system also integrates an external data interface for acquiring weather forecasts, electricity price information, and air quality index data with a time resolution of at least one hour. The actuators include a variable frequency fan, a damper actuator, and a surface cooler valve controller. The functional layer is implemented by program modules running on the controller, including a data sensing and processing module, a deep learning prediction module, a fuzzy logic control module, and a global optimization module.

[0019] The data sensing and processing module is configured to acquire indoor and outdoor environmental parameters from the sensor array and perform preprocessing operations on the data, including denoising, filtering, and normalization. The deep learning prediction module is configured to receive the preprocessed data from the data sensing and processing module and build a time series prediction model based on an LSTM network. Taking historical environmental data, equipment operating status, and external environmental data as input, it outputs the trend of environmental parameters and predicted system energy consumption values ​​for a future period. The time window length N for the historical environmental data ranges from 12 to 168, and the future period M ranges from 1 to 24.

[0020] The fuzzy logic control module is configured to receive real-time environmental data from the data perception and processing module and predicted data from the deep learning prediction module. It converts the data into fuzzy linguistic variables by comparing them with a preset quantization range, and then performs inference based on a fuzzy rule base to generate preliminary control commands. The global optimization module is configured to receive predicted data from the deep learning prediction module, perform multi-objective optimization using an optimization algorithm, generate a globally optimal operating strategy, and dynamically adjust the setpoints or rule parameters of the fuzzy logic control module based on this strategy. The controller is configured to send the control commands generated by the fuzzy logic control module to the actuators to drive the ventilation equipment, wherein the speed setpoint of the variable frequency fan is adjustable from 20% to 100% of the rated speed.

[0021] In a preferred embodiment, the optimization algorithm employed by the global optimization module is a genetic algorithm, with a population size set to 50-200, a crossover probability set to 0.7-0.9, and a mutation probability set to 0.01-0.1. In another preferred embodiment, the global optimization module employs a reinforcement learning algorithm, with the reward function of the reinforcement learning agent incorporating energy cost, thermal comfort indicators, and indoor air quality indicators. The weights of the indicators in the reward function can be dynamically adjusted in response to external signals, with the weight of the energy cost indicator being increased in proportion to the increase in electricity price upon receiving a demand response signal from the power grid.

[0022] The system can also include a digital twin module configured to construct a high-fidelity virtual model of the ventilation system and the building in which it is located. The operation strategy generated by the global optimization module is first simulated and iterated in the digital twin module before being deployed to the physical system for execution. The fuzzy logic control module can employ an adaptive neuro-fuzzy inference system (ANFIS) configured to self-learn and adjust the membership functions and fuzzy rule parameters based on historical operation data of the system.

[0023] The system also includes a human-machine interaction module for receiving subjective comfort feedback from users. The global optimization module and / or the fuzzy logic control module are further configured to dynamically adjust comfort-related parameters or set values within them based on the feedback data.

[0024] The system achieves intelligent regulation and energy optimization of the building ventilation system through the coordinated operation of the above hardware and software. The modules cooperate with each other, with the data perception and processing module providing accurate data basis, the deep learning prediction module providing forward-looking trend prediction, the global optimization module formulating the optimal operation strategy, and the fuzzy logic control module achieving precise real-time control, thereby ensuring that the system always maintains an efficient and energy-saving operation state under changing environmental conditions.

[0025] The system was applied to a 5000 square meter office building, and after one cooling season of operation test, compared with the traditional constant air volume (CAV) system, the following effects were achieved: Energy saving rate: the energy consumption of the ventilation system was reduced by 35%.

[0026] Comfort: the proportion of time within the set value ± 0.5℃ of indoor temperature was increased from 78% to 95%.

[0027] Air quality: the proportion of time with CO2 concentration exceeding the standard (>1000ppm) was reduced to less than 1%.

[0028] Demand response: successfully responded to two peak demands of the power grid, with an average load reduction of 25kW, and obtained electricity cost savings.

[0029] While embodiments of the application have been shown and described, it is to be understood that the embodiments described are merely exemplary of the principles and application of the present application. Numerous modifications and adaptions can be effected without departing from the spirit and scope of the present application, which is not limited to the exact construction and arrangement described. It is intended, therefore, to cover all modifications and adaptions that fall within the scope of the claims and their equivalents.

Claims

1. A smart energy efficient building ventilation system that adapts to environmental changes, characterized by, The application relates to a ventilation system, comprising: a hardware layer and a function layer, wherein the hardware layer comprises a sensor group, an actuator and a controller in communication connection with the sensor group and the actuator; the function layer is realized by program modules run by the controller, and comprises: a data sensing and processing module configured to acquire indoor and outdoor environment parameters from the sensor group and perform preprocessing including denoising, filtering and normalization on the data; a deep learning prediction module configured to receive preprocessed data from the data sensing and processing module and establish a time series prediction model based on an LSTM network, wherein historical environment data, device running state and external environment data are input, and environment parameter trends and system energy consumption prediction values in a future period of time are output; a fuzzy logic control module configured to receive real-time environment data from the data sensing and processing module and receive prediction data from the deep learning prediction module, compare the data with preset quantization ranges to convert the data into fuzzy language variables, and generate preliminary control instructions according to a fuzzy rule base; a global optimization module configured to receive prediction data from the deep learning prediction module, perform multi-objective optimization solving through an optimization algorithm, generate a globally optimal running strategy, and dynamically adjust the set value or rule parameter of the fuzzy logic control module based on the strategy; wherein the controller is configured to send the control instructions generated by the fuzzy logic control module to the actuator to drive the ventilation equipment to run.

2. The system of claim 1, wherein, The time window length N of the historical environment data is 12-168, and the future period of time M is 1-24.

3. The system of claim 1, wherein, The optimization algorithm adopted by the global optimization module is a genetic algorithm, the population size of which is set to 50-200, the crossover probability is set to 0.7-0.9, and the mutation probability is set to 0.01-0.

1.

4. The system of claim 1, wherein, The optimization algorithm adopted by the global optimization module is a reinforcement learning algorithm; a reward function of the reinforcement learning agent comprehensively considers energy consumption cost, thermal comfort index and indoor air quality index.

5. The system of claim 4, wherein, The weights of the indexes in the reward function can be dynamically adjusted in response to external signals; when a power grid demand response signal is received, the weight of the energy consumption cost index is increased according to the price increase ratio.

6. The system of claim 1, wherein, The application further comprises a digital twin module configured to construct a high-fidelity virtual model of the ventilation system and a building where the ventilation system is located; the running strategy generated by the global optimization module is first simulated, verified and iterated in the digital twin module and then deployed to the physical system for execution.

7. The system of claim 1, wherein, The fuzzy logic control module is an adaptive neuro-fuzzy inference system (ANFIS) configured to self-learn and adjust membership functions and fuzzy rule parameters according to system historical running data.

8. The system of claim 1, wherein, The sensor group comprises an indoor carbon dioxide concentration sensor, a temperature and humidity sensor, a personnel counter and an outdoor temperature and humidity sensor; the system further integrates an external data interface for acquiring weather forecasts, electricity price information and air quality index data with a time resolution of not less than 1 hour.

9. The system of claim 1, wherein, The actuators include variable frequency blowers, air valve actuators and table cooler valve controllers; the control instructions include a rotational speed setting value of the variable frequency blower, an opening degree instruction of each regional air valve and an opening degree instruction of the table cooler valve, wherein the rotational speed setting value of the variable frequency blower is adjusted in a range of 20% to 100% of a rated rotational speed.

10. The system of claim 1, wherein, A human-computer interaction module is further included for receiving subjective comfort feedback of a user; the global optimization module and / or the fuzzy logic control module are further configured to dynamically adjust comfort-related parameters or setting values in them based on the feedback data.

Images