Constant-temperature and constant-humidity adjusting method for multi-area cooperative control and related device

Abstract

The invention discloses a multi-area cooperative control constant temperature and humidity adjusting method and a related device, and the method comprises the steps: dividing a to-be-adjusted area into a plurality of sub-areas, deploying a sensor in each sub-area, and collecting the sensing data of the corresponding sub-area in real time; the collected sensing data is preprocessed; constructing an edge side distributed thermal load prediction model, inputting the preprocessed sensing data into the edge side distributed thermal load prediction model, and obtaining regional thermal load prediction data of each sub-region; performing gradient strategy analysis on the regional thermal load prediction data to generate a constant-temperature and constant-humidity cooperative adjustment instruction; and performing constant-temperature and constant-humidity adjustment on a to-be-adjusted area based on the constant-temperature and constant-humidity cooperative adjustment instruction The problem of collaborative delay of constant-temperature and constant-humidity adjustment among multiple areas is solved, the constant-temperature and constant-humidity adjustment efficiency of the to-be-adjusted area is effectively improved, energy consumption is optimized, and a better constant-temperature and constant-humidity adjustment effect is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of constant temperature and humidity regulation, and particularly to a constant temperature and humidity regulation method and related device for multi-region collaborative control. Background Art

[0002] Since large commercial complexes are usually large-area enclosed spaces, it is necessary to install constant temperature and humidity air conditioning equipment to maintain a constant environmental temperature and humidity in order to bring people a good entertainment experience. At the same time, due to the existence of multiple regions with different functions, the population flow in each region fluctuates greatly and the environmental load changes frequently. The traditional centralized PID control algorithm, which adjusts the temperature and humidity of the entire region through a single controller, has the problem of multi-region collaborative delay, resulting in temperature and humidity oscillation between different regions, low adjustment efficiency, poor effect, and affecting the user experience. Summary of the Invention

[0003] The purpose of the present invention is to overcome the deficiencies of the prior art. The present invention provides a constant temperature and humidity regulation method and related device for multi-region collaborative control, solves the problem of collaborative delay in constant temperature and humidity regulation between multiple regions, effectively improves the constant temperature and humidity regulation efficiency of the area to be regulated, optimizes energy consumption, and achieves a better constant temperature and humidity regulation effect.

[0004] The present invention provides a constant temperature and humidity regulation method for multi-region collaborative control, and the method includes:

[0005] Dividing the area to be regulated into several sub-regions, deploying sensors in each sub-region, and collecting sensing data of the corresponding sub-region in real time based on the sensors;

[0006] Preprocessing the collected sensing data to obtain preprocessed sensing data;

[0007] Constructing an edge-side distributed heat load prediction model, inputting the preprocessed sensing data into the edge-side distributed heat load prediction model, and obtaining regional heat load prediction data of each sub-region;

[0008] Performing gradient strategic analysis on the regional heat load prediction data to generate a constant temperature and humidity collaborative regulation instruction;

[0009] Performing constant temperature and humidity regulation on the area to be regulated based on the constant temperature and humidity collaborative regulation instruction.

[0010] Further, the collecting sensing data of the corresponding sub-region in real time based on the sensors includes:

[0011] Collecting environmental temperature and humidity data, equipment power data, and population density data of the corresponding sub-region in real time based on the sensors.

[0012] Further, the preprocessing of the collected sensing data to obtain the preprocessed sensing data includes:

[0013] Performing filtering denoising and feature extraction on the collected sensing data to generate a heat load prediction input vector containing spatio-temporal distribution.

[0014] Further, the construction of the edge-side distributed heat load prediction model and the input of the preprocessed sensing data into the edge-side distributed heat load prediction model to obtain the regional heat load prediction data for each sub-region include:

[0015] Setting regional edge nodes in each sub-region and constructing an edge-side distributed heat load prediction model based on the regional edge nodes combined with the LSTM neural network;

[0016] Setting and allocating priority weights for each sub-region based on the heat load prediction input vector;

[0017] Inputting the preprocessed sensing data into the edge-side distributed heat load prediction model and performing weighted average processing according to the priority weights of each sub-region, and outputting the regional heat load prediction data for each sub-region.

[0018] Further, the setting and allocation of priority weights for each sub-region based on the heat load prediction input vector include:

[0019] Collecting historical adjustment data and setting and allocating priority weights for each sub-region based on the heat load prediction input vector combined with the historical adjustment data.

[0020] Further, the gradient strategic analysis of the regional heat load prediction data to generate a constant temperature and humidity collaborative adjustment instruction includes:

[0021] Inputting the regional heat load prediction data into a double-delay deep deterministic policy gradient model for reinforcement learning agent to generate an initial constant temperature and humidity collaborative adjustment instruction;

[0022] Collecting the zero-drift error amount of the sensing data collected by the sensor based on the Kalman filter algorithm;

[0023] Analyzing the unmodeled mutation factors in the sensing data collected by the sensor based on the LSTM neural network and outputting an adjustment perturbation compensation amount;

[0024] Performing feedforward compensation on the initial constant temperature and humidity collaborative adjustment instruction based on the zero-offset error amount and the adjustment perturbation compensation amount to generate a constant temperature and humidity collaborative adjustment instruction.

[0025] Further, the constant temperature and humidity adjustment of the area to be adjusted based on the constant temperature and humidity collaborative adjustment instruction includes:

[0026] Perform convergence verification on the constant temperature and humidity collaborative regulation instruction.

[0027] The present invention also provides a constant temperature and humidity regulation system for multi-region collaborative control, which is used to implement the above-mentioned constant temperature and humidity regulation method for multi-region collaborative control. The system includes:

[0028] A data acquisition module, which is used to divide the area to be regulated into several sub-areas, deploy sensors in each sub-area, and collect sensing data of the corresponding sub-area in real time based on the sensors;

[0029] A data preprocessing module, which is used to preprocess the collected sensing data to obtain the preprocessed sensing data;

[0030] A regional heat load prediction module, which is used to build an edge-side distributed heat load prediction model, input the preprocessed sensing data into the edge-side distributed heat load prediction model, and obtain the regional heat load prediction data of each sub-area;

[0031] A collaborative regulation instruction generation module, which is used to perform gradient strategic analysis on the regional heat load prediction data to generate a constant temperature and humidity collaborative regulation instruction;

[0032] A constant temperature and humidity regulation module, which is used to perform constant temperature and humidity regulation on the area to be regulated based on the constant temperature and humidity collaborative regulation instruction.

[0033] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. The processor executes the computer program to implement the above-mentioned constant temperature and humidity regulation method for multi-region collaborative control.

[0034] The present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the above-mentioned constant temperature and humidity regulation method for multi-region collaborative control.

[0035] The present invention provides a constant temperature and humidity regulation method and related device for multi-region collaborative control. By adopting an edge-side distributed heat load prediction model constructed based on regional edge nodes, the regional heat load prediction data of each sub-region in the region to be regulated is obtained, and then gradient strategic analysis is carried out to generate a collaborative regulation instruction, solving the problem of collaborative delay in constant temperature and humidity regulation between multiple regions, avoiding the problem of temperature and humidity oscillation between sub-regions, effectively improving the constant temperature and humidity regulation efficiency of the region to be regulated, optimizing energy consumption, and achieving a better constant temperature and humidity regulation effect. Brief Description of the Drawings

[0036] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0037] Figure 1 It is a flowchart of the constant temperature and humidity regulation method for multi-region collaborative control in Embodiment 1 of the present invention;

[0038] Figure 2 It is a flowchart of obtaining the regional heat load prediction data of each sub-region in Embodiment 1 of the present invention;

[0039] Figure 3 It is a flowchart of gradient strategic analysis of the regional heat load prediction data in Embodiment 1 of the present invention;

[0040] Figure 4 It is an architecture diagram of the constant temperature and humidity regulation system for multi-region collaborative control in Embodiment 2 of the present invention. Detailed Description of the Embodiments

[0041] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0042] In the present invention, it should be understood that terms such as "including" or "having" are intended to indicate the existence of the features, numbers, steps, actions, components, parts, or combinations thereof disclosed in this specification, and do not intend to exclude the possibility of the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0043] It should be further noted that, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0044] Embodiment 1

[0045] Embodiment 1 of the present invention provides a constant temperature and humidity regulation method for multi-region collaborative control. The method includes: dividing the area to be regulated into several sub-areas, deploying sensors in each sub-area, and based on the sensors, collecting sensing data of the corresponding sub-area in real time; preprocessing the collected sensing data to obtain the preprocessed sensing data; constructing an edge-side distributed heat load prediction model, inputting the preprocessed sensing data into the edge-side distributed heat load prediction model to obtain the regional heat load prediction data of each sub-area; performing gradient strategic analysis on the regional heat load prediction data to generate a constant temperature and humidity collaborative regulation instruction; and performing constant temperature and humidity regulation on the area to be regulated based on the constant temperature and humidity collaborative regulation instruction.

[0046] In an alternative implementation manner of this embodiment, as Figure 1 shown, Figure 1 shows the flowchart of the constant temperature and humidity regulation method for multi-region collaborative control in Embodiment 1 of the present invention, including the following steps:

[0047] S101. Divide the area to be regulated into several sub-areas, deploy sensors in each sub-area, and based on the sensors, collect sensing data of the corresponding sub-area in real time;

[0048] In an alternative implementation manner of this embodiment, the area to be regulated is divided into several sub-areas, and the division basis may include area size, regional function, etc. In this embodiment, the main application scenarios are large commercial complexes, such as shopping malls, convention centers, etc. There are usually areas with different functions, including fresh food areas, dining areas, daily necessities areas, etc. The number of people in each area is different, and the environmental temperature and humidity to be maintained are also different. Therefore, the division basis adopted is regional function.

[0049] In an alternative implementation manner of this embodiment, sensors are deployed in each sub-area, and the types of sensors include temperature sensors, humidity sensors, power sensors, and infrared sensors.

[0050] In an alternative implementation manner of this embodiment, the real-time collection of sensing data of the corresponding sub-area based on the sensors includes: real-time collection of environmental temperature and humidity data, equipment power data, and people flow density data of the corresponding sub-area based on the sensors.

[0051] Specifically, based on the temperature sensor, the ambient temperature data of the corresponding sub-region is collected in real time. Based on the humidity sensor, the ambient humidity data of the corresponding sub-region is collected in real time. Based on the power sensor, the device power data of the corresponding sub-region is collected in real time, and based on the infrared sensor, the pedestrian flow density data of the corresponding sub-region is collected in real time.

[0052] S102. Preprocess the collected sensing data to obtain the preprocessed sensing data.

[0053] In an optional implementation manner of this embodiment, the collected sensing data is subjected to filtering and denoising and feature extraction processing to generate a thermal load prediction input vector including spatio-temporal distribution.

[0054] In an optional implementation manner of this embodiment, first, the ambient temperature and humidity data, device power data, and pedestrian flow density data of the corresponding sub-region collected by each sensor are subjected to filtering and denoising processing.

[0055] Specifically, let the collected original sensing data be X(t) = {x1(t), x2(t), x3(t)}. The original sensing data is subjected to moving average filtering and denoising processing, and the calculation formula includes:

[0056]

[0057] In the formula, is the sensing data after moving average filtering and denoising. x1(t) is the ambient temperature and humidity data, x2(t) is the device power data, x3(t) is the pedestrian flow density data, N is the window size, taking the value of 30, corresponding to a 5-minute sliding time window with a sampling frequency of 1 Hz.

[0058] Furthermore, the sensing data after moving average filtering and denoising is subjected to discrete wavelet transform processing for high-frequency noise suppression, and the calculation formula includes:

[0059]

[0060] In the formula, is the sensing data after discrete wavelet transform processing. ψ(t) is the mother wavelet function, j is the decomposition scale, taking the value of 3, k is the translation parameter, is the normalization factor.

[0061] In an optional implementation manner of this embodiment, feature extraction processing is performed on the sensing data after filtering and denoising.

[0062] Specifically, after performing time series encoding on the ambient temperature and humidity data, device power data, and pedestrian flow density data in the sensing data after filtering and denoising respectively, feature vector weighting is performed based on the UWB positioning coordinates, and then a thermal load prediction input vector including spatio-temporal distribution is constructed and generated.

[0063] In an alternative implementation of this embodiment, a time-dependent feature vector f1 is extracted from the environmental temperature and humidity data based on the LSTM network, a power spectrum feature vector f2 is extracted from the device power data based on the short-time Fourier transform, and a spatial distribution density feature vector f3 is extracted from the pedestrian flow density data based on the DBSCAN clustering algorithm.

[0064] Furthermore, the three feature vectors are weighted based on the UWB positioning coordinates (u, v), and the calculation formula includes:

[0065] F(u, v) = α·f1 + β·f2 + γ·f3

[0066] In the formula, F(u, v) is the spatio-temporal feature vector, and α, β, and γ are position weight coefficients, which are dynamically adjusted according to the functional type of the sub-region.

[0067] Furthermore, the spatio-temporal feature vector is normalized to obtain the final heat load prediction input vector F′(u, v) including spatio-temporal distribution.

[0068] S103. Construct an edge-side distributed heat load prediction model, input the preprocessed sensing data into the edge-side distributed heat load prediction model, and obtain the regional heat load prediction data of each sub-region;

[0069] In an alternative implementation of this embodiment, as Figure 2 shown, Figure 2 shows the flowchart of obtaining the regional heat load prediction data of each sub-region in the first embodiment of the present invention, including the following steps:

[0070] S201. Set regional edge nodes in each sub-region, and construct an edge-side distributed heat load prediction model based on the regional edge nodes combined with the LSTM neural network;

[0071] In an alternative implementation of this embodiment, regional edge nodes are deployed at the boundaries of the divided sub-regions, and an edge-side distributed heat load prediction model is constructed based on each regional edge node combined with the LSTM neural network. The edge-side distributed heat load prediction model includes a position model of each regional edge node, an LSTM prediction engine, and a learning controller. The LSTM prediction engine is a multi-layer LSTM network constructed based on gated recurrent units, and the input dimension is environmental temperature and humidity, device power, and pedestrian flow density. The learning controller is a model parameter controller that aggregates adjacent sub-regions based on differential privacy technology.

[0072] S202. Set and assign priority weights for each sub-region based on the heat load prediction input vector;

[0073] In an alternative implementation of this embodiment, historical adjustment data is collected, and allocation priority weights are set for each sub-region based on the heat load prediction input vector in combination with the historical adjustment data.

[0074] Specifically, the heat load prediction input vector can be corrected here by adding historical adjustment data, which specifically includes:

[0075] Concatenate the heat load prediction input vector F′(u, v) containing spatio-temporal distribution with the historical adjustment data H(t) to obtain the corrected heat load prediction input vector Q i (t), Q i (t) = [F ′ (u, v); H(t)].

[0076] In an alternative implementation of this embodiment, allocation priority weights are set for each sub-region based on the corrected heat load prediction input vector Q i (t), and the formula includes:

[0077]

[0078] In the formula, ω i (t) is the priority weight, ΔQ i (t) is the difference between the corrected heat load prediction input vector Q i (t) and the heat load in the same historical period, and δ is the adjustment factor.

[0079] S203. Input the preprocessed sensing data into the edge-side distributed heat load prediction model, and perform weighted average processing according to the priority weights of each sub-region, and output the regional heat load prediction data of each sub-region.

[0080] In an alternative implementation of this embodiment, input the preprocessed sensing data into the edge-side distributed heat load prediction model to generate the original prediction value P i (t) of each sub-region, and perform weighted average processing on the original prediction value P i (t) according to the priority weights of each sub-region, and output the regional heat load prediction data of each sub-region. The calculation formula includes:

[0081]

[0082] In the formula, P ′ (t) is the regional heat load prediction data of each sub-region.

[0083] S104. Perform gradient strategic analysis on the regional heat load prediction data to generate a constant temperature and humidity collaborative adjustment instruction;

[0084] In an alternative implementation of this embodiment, as Figure 3 shown, Figure 3 The flowchart of gradient strategic analysis of the regional heat load prediction data in the first embodiment of the present invention is shown, including the following steps:

[0085] S301. Input the regional heat load prediction data into a Twin Delayed Deep Deterministic Policy Gradient (TD3) model for reinforcement learning agent to generate an initial instruction for coordinated regulation of constant temperature and humidity;

[0086] In an alternative implementation of this embodiment, the regional heat load prediction data P ′ (t) of each sub-region generated in step S103 is input into a Twin Delayed Deep Deterministic Policy Gradient (TD3) model for reinforcement learning agent to generate an initial instruction for coordinated regulation of constant temperature and humidity.

[0087] Specifically, the Twin Delayed Deep Deterministic Policy Gradient (TD3) model is a reinforcement learning algorithm for continuous action spaces. It uses two independent Critic networks, delays policy updates, smooths the target policy, has strong stability, robustness, and more balanced optimization, and is suitable for continuous control scenarios. In this embodiment, since continuous coordinated regulation of constant temperature and humidity is required for each sub-region of the area to be regulated, the Twin Delayed Deep Deterministic Policy Gradient (TD3) model is used here, and the generated initial instruction for coordinated regulation of constant temperature and humidity has higher accuracy and better coordination, effectively improving the accuracy and efficiency of coordinated regulation.

[0088] Furthermore, the generated initial instruction for coordinated regulation of constant temperature and humidity includes state space regulation, action space regulation, and a reward function. Among them, state space regulation includes temperature and humidity adjustment amount, energy consumption cost adjustment amount, and load fluctuation adjustment amount, and action space regulation includes start-stop regulation of chillers, speed regulation of fans, opening regulation of humidifying valves, etc.

[0089] S302. Collect the zero-drift error amount of the sensing data collected by the sensor based on the Kalman filter algorithm;

[0090] In an alternative implementation of this embodiment, collecting the zero-drift error amount of the sensing data collected by the sensor based on the Kalman filter algorithm is specifically to calculate the reading jump rate z k of the sensing data by building a process noise covariance matrix and a measurement noise covariance matrix. When the reading jump rate is greater than a certain threshold, drift correction is triggered, and a zero-drift error amount ε k is generated.

[0091] S303. Analyze the unmodeled mutation factors in the sensing data collected by the sensor based on the LSTM neural network, and output the adjustment disturbance compensation amount;

[0092] In an optional implementation manner of this embodiment, input the unmodeled mutation factors in the sensing data collected by the sensor into the LSTM neural network to predict the unmodeled dynamic disturbance Δu. The calculation formula includes:

[0093] Δu = LSTM φ (z k ) = φ(W1z k +W2h k-1 +W3)

[0094] In the formula, Δu is the unmodeled dynamic disturbance, z k is the reading jump rate, h k-1 is the hidden layer state, W1, W2, and W3 are variable parameters of the dynamic disturbance, and φ is the activation function.

[0095] Furthermore, generate the adjustment disturbance compensation amount u based on the unmodeled dynamic disturbance Δu. The calculation formula includes:

[0096] u = Δu·η

[0097] In the formula, u is the adjustment disturbance compensation amount, Δu is the unmodeled dynamic disturbance, and η is the compensation coefficient.

[0098] S304. Perform feedforward compensation on the initial instruction for constant temperature and humidity collaborative regulation based on the zero offset error amount and the adjustment disturbance compensation amount to generate an instruction for constant temperature and humidity collaborative regulation.

[0099] In an optional implementation manner of this embodiment, superimpose the zero offset error amount ε k and the adjustment disturbance compensation amount u into the initial instruction for constant temperature and humidity collaborative regulation in a feedforward compensation manner to generate an instruction for constant temperature and humidity collaborative regulation.

[0100] S105. Perform constant temperature and humidity regulation on the area to be regulated based on the instruction for constant temperature and humidity collaborative regulation.

[0101] In an optional implementation manner of this embodiment, perform constant temperature and humidity regulation on the corresponding sub-area of the area to be regulated based on the instruction for constant temperature and humidity collaborative regulation generated in step S104.

[0102] In an optional implementation manner of this embodiment, perform convergence verification on the instruction for constant temperature and humidity collaborative regulation.

[0103] Specifically, here, by defining multi-objective convergence verification metrics, including stability metrics, control accuracy metrics, and energy consumption constraint metrics, the convergence of the constant temperature and humidity collaborative regulation instruction is verified. When it is determined that the convergence verification of the constant temperature and humidity collaborative regulation instruction passes, it means that the rationality of implementing the constant temperature and humidity collaborative regulation instruction is verified, and then the specific constant temperature and humidity regulation can be carried out.

[0104] In summary, Embodiment 1 of the present invention provides a constant temperature and humidity regulation method for multi-region collaborative control. By adopting an edge-side distributed heat load prediction model constructed based on regional edge nodes, the regional heat load prediction data of each sub-region in the region to be regulated is obtained, and then gradient strategic analysis is carried out to generate a collaborative regulation instruction, solving the problem of collaborative delay in constant temperature and humidity regulation between multiple regions, avoiding the problem of temperature and humidity oscillation between sub-regions, effectively improving the constant temperature and humidity regulation efficiency of the region to be regulated, optimizing energy consumption, and achieving a better constant temperature and humidity regulation effect.

[0105] Embodiment 2

[0106] Embodiment 2 of the present invention provides a constant temperature and humidity regulation system for multi-region collaborative control. The constant temperature and humidity regulation system for multi-region collaborative control is used to implement the above-mentioned constant temperature and humidity regulation method for multi-region collaborative control. The system includes a data acquisition module, a data preprocessing module, a regional heat load prediction module, a collaborative regulation instruction generation module, and a constant temperature and humidity regulation module.

[0107] In an optional implementation manner of this embodiment, as Figure 4 shown, Figure 4 shows the architecture diagram of the constant temperature and humidity regulation system for multi-region collaborative control in Embodiment 2 of the present invention, including the following modules:

[0108] Data acquisition module 10, the data acquisition module 10 is used to divide the region to be regulated into several sub-regions, deploy sensors in each sub-region, and based on the sensors, collect sensing data of the corresponding sub-region in real time;

[0109] In an optional implementation manner of this embodiment, the real-time collection of sensing data of the corresponding sub-region based on the sensors includes:

[0110] Based on the sensors, collect environmental temperature and humidity data, equipment power data, and population density data of the corresponding sub-region in real time.

[0111] Data preprocessing module 20, the data preprocessing module 20 is used to preprocess the collected sensing data to obtain the preprocessed sensing data;

[0112] In an alternative implementation of this embodiment, the preprocessing of the collected sensing data to obtain the preprocessed sensing data includes:

[0113] Performing filtering denoising and feature extraction processing on the collected sensing data to generate a heat load prediction input vector including spatio-temporal distribution.

[0114] The regional heat load prediction module 30 is configured to construct an edge-side distributed heat load prediction model, input the preprocessed sensing data into the edge-side distributed heat load prediction model, and obtain the regional heat load prediction data for each sub-region;

[0115] In an alternative implementation of this embodiment, the constructing the edge-side distributed heat load prediction model, inputting the preprocessed sensing data into the edge-side distributed heat load prediction model, and obtaining the regional heat load prediction data for each sub-region includes:

[0116] Setting regional edge nodes in each sub-region, and constructing an edge-side distributed heat load prediction model based on the regional edge nodes in combination with an LSTM neural network;

[0117] Setting and allocating priority weights for each sub-region based on the heat load prediction input vector;

[0118] Inputting the preprocessed sensing data into the edge-side distributed heat load prediction model, and performing weighted average processing according to the priority weights of each sub-region to output the regional heat load prediction data for each sub-region.

[0119] In an alternative implementation of this embodiment, the setting and allocating priority weights for each sub-region based on the heat load prediction input vector includes:

[0120] Collecting historical adjustment data, and setting and allocating priority weights for each sub-region based on the heat load prediction input vector in combination with the historical adjustment data.

[0121] The collaborative adjustment instruction generation module 40 is configured to perform gradient strategic analysis on the regional heat load prediction data to generate a constant temperature and humidity collaborative adjustment instruction;

[0122] In an alternative implementation of this embodiment, the performing gradient strategic analysis on the regional heat load prediction data to generate a constant temperature and humidity collaborative adjustment instruction includes:

[0123] Inputting the regional heat load prediction data into a double-delay deep deterministic policy gradient model for reinforcement learning agent to generate a constant temperature and humidity collaborative adjustment initial instruction;

[0124] Collect the zero - drift error amount of the sensing data collected by the sensor based on the Kalman filter algorithm;

[0125] Analyze the unmodeled mutation factors in the sensing data collected by the sensor based on the LSTM neural network, and output the adjustment disturbance compensation amount;

[0126] Perform feed - forward compensation on the initial instruction for constant temperature and humidity collaborative regulation based on the zero - offset error amount and the adjustment disturbance compensation amount, and generate an instruction for constant temperature and humidity collaborative regulation.

[0127] In an alternative implementation of this embodiment, the constant temperature and humidity regulation of the area to be regulated based on the instruction for constant temperature and humidity collaborative regulation includes:

[0128] Verify the convergence of the instruction for constant temperature and humidity collaborative regulation.

[0129] A constant temperature and humidity regulation module 50, which is used to perform constant temperature and humidity regulation on the area to be regulated based on the instruction for constant temperature and humidity collaborative regulation.

[0130] In summary, Embodiment 2 of the present invention provides a constant temperature and humidity regulation system for multi - area collaborative control, which is used to implement the constant temperature and humidity regulation method for multi - area collaborative control in Embodiment 1. By adopting an edge - side distributed heat load prediction model constructed based on regional edge nodes, the regional heat load prediction data of each sub - area in the area to be regulated is obtained, and then gradient strategic analysis is performed to generate a collaborative regulation instruction, solving the problem of collaborative delay in constant temperature and humidity regulation between multiple areas, avoiding the problem of temperature and humidity oscillation between sub - areas, effectively improving the constant temperature and humidity regulation efficiency of the area to be regulated, optimizing energy consumption, and achieving a better constant temperature and humidity regulation effect.

[0131] Embodiment 3

[0132] Embodiment 3 of the present invention provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. The processor executes the computer program to implement the constant temperature and humidity regulation method for multi - area collaborative control described in Embodiment 1.

[0133] In summary, Embodiment 3 of the present invention provides an electronic device for implementing the multi-region collaborative control constant temperature and humidity regulation method described in Embodiment 1. By adopting an edge-side distributed heat load prediction model constructed based on regional edge nodes, the regional heat load prediction data of each sub-region in the region to be regulated is obtained, and then gradient strategic analysis is performed to generate a collaborative regulation instruction, solving the problem of collaborative delay in constant temperature and humidity regulation between multiple regions, avoiding the problem of temperature and humidity oscillation between sub-regions, effectively improving the constant temperature and humidity regulation efficiency of the region to be regulated, optimizing energy consumption, and achieving a better constant temperature and humidity regulation effect.

[0134] Embodiment 4

[0135] Embodiment 4 of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the multi-region collaborative control constant temperature and humidity regulation method described in Embodiment 1.

[0136] In summary, Embodiment 4 of the present invention provides a computer-readable storage medium for executing the multi-region collaborative control constant temperature and humidity regulation method described in Embodiment 1. By adopting an edge-side distributed heat load prediction model constructed based on regional edge nodes, the regional heat load prediction data of each sub-region in the region to be regulated is obtained, and then gradient strategic analysis is performed to generate a collaborative regulation instruction, solving the problem of collaborative delay in constant temperature and humidity regulation between multiple regions, avoiding the problem of temperature and humidity oscillation between sub-regions, effectively improving the constant temperature and humidity regulation efficiency of the region to be regulated, optimizing energy consumption, and achieving a better constant temperature and humidity regulation effect.

[0137] The above has introduced in detail a multi-region collaborative control constant temperature and humidity regulation method and related devices provided by the present invention. Those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing relevant hardware through a program, and this program can be stored in a computer-readable storage medium. The storage medium can include: read-only memory (ROM, Read Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disc, etc.

[0138] In addition, the above embodiments of the present invention have been introduced in detail. Specific examples are used in this article to elaborate on the principle and implementation manner of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation on the present invention.

Claims

1. A constant temperature and humidity regulation method with multi-region collaborative control, characterized in that, The method includes: Dividing the area to be adjusted into several sub - areas, deploying sensors in each sub - area, and collecting sensing data of the corresponding sub - area in real time based on the sensors; Pre - processing the collected sensing data to obtain the pre - processed sensing data; Constructing an edge - side distributed heat load prediction model, inputting the pre - processed sensing data into the edge - side distributed heat load prediction model, and obtaining the regional heat load prediction data of each sub - area; Performing gradient strategic analysis on the regional heat load prediction data to generate a constant temperature and humidity collaborative regulation instruction; Performing constant temperature and humidity regulation on the area to be adjusted based on the constant temperature and humidity collaborative regulation instruction.

2. The multi-zone collaborative control method for constant temperature and humidity regulation according to claim 1, wherein The collecting sensing data of the corresponding sub - area in real time based on the sensors includes: Collecting the ambient temperature and humidity data, equipment power data, and population density data of the corresponding sub - area in real time based on the sensors.

3. The multi-region collaborative control method for constant temperature and humidity regulation according to claim 1, characterized in that The pre - processing the collected sensing data to obtain the pre - processed sensing data includes: Performing filtering and denoising and feature extraction on the collected sensing data to generate a heat load prediction input vector including spatio - temporal distribution.

4. The multi-zone collaborative control method for constant temperature and humidity regulation according to claim 3, characterized in that The constructing an edge - side distributed heat load prediction model, inputting the pre - processed sensing data into the edge - side distributed heat load prediction model, and obtaining the regional heat load prediction data of each sub - area includes: Setting regional edge nodes in each sub - area, and constructing an edge - side distributed heat load prediction model based on the regional edge nodes combined with the LSTM neural network; Setting and allocating priority weights for each sub - area based on the heat load prediction input vector; Inputting the pre - processed sensing data into the edge - side distributed heat load prediction model, and performing weighted average processing according to the priority weights of each sub - area, and outputting the regional heat load prediction data of each sub - area.

5. The multi-region collaborative control method for constant temperature and humidity regulation according to claim 4, characterized in that, The setting and allocating priority weights for each sub - area based on the heat load prediction input vector includes: Collecting historical regulation data, and setting and allocating priority weights for each sub - area based on the heat load prediction input vector combined with the historical regulation data.

6. The constant temperature and humidity regulation method with multi-region collaborative control according to claim 1, characterized in that, The performing gradient strategic analysis on the regional heat load prediction data to generate a constant temperature and humidity collaborative regulation instruction includes: Inputting the regional heat load prediction data into a double - delay deep deterministic policy gradient model for reinforcement learning agent to generate an initial constant temperature and humidity collaborative regulation instruction; Collecting the zero - drift error amount of the sensing data collected by the sensors based on the Kalman filter algorithm; Analyzing the unmodeled mutation factors in the sensing data collected by the sensors based on the LSTM neural network, and outputting an adjustment perturbation compensation amount; Performing feed - forward compensation on the initial constant temperature and humidity collaborative regulation instruction based on the zero - offset error amount and the adjustment perturbation compensation amount to generate a constant temperature and humidity collaborative regulation instruction.

7. The method for constant temperature and humidity regulation with multi - region collaborative control according to claim 6, characterized in that, The performing constant temperature and humidity regulation on the area to be adjusted based on the constant temperature and humidity collaborative regulation instruction includes: Performing convergence verification on the constant temperature and humidity collaborative regulation instruction.

8. A constant temperature and humidity regulation system with multi-region collaborative control, characterized in that, The constant temperature and humidity regulation system for multi - area collaborative control is used to implement the constant temperature and humidity regulation method for multi - area collaborative control according to any one of claims 1 - 7. The system includes: A data acquisition module, which is used to divide the area to be adjusted into several sub-areas, deploy sensors in each sub-area, and collect sensing data of the corresponding sub-area in real time based on the sensors; A data preprocessing module, which is used to preprocess the collected sensing data to obtain the preprocessed sensing data; A regional heat load prediction module, which is used to build an edge-side distributed heat load prediction model, input the preprocessed sensing data into the edge-side distributed heat load prediction model, and obtain the regional heat load prediction data of each sub-area; A collaborative adjustment instruction generation module, which is used to perform gradient strategic analysis on the regional heat load prediction data to generate a constant temperature and humidity collaborative adjustment instruction; A constant temperature and humidity adjustment module, which is used to perform constant temperature and humidity adjustment on the area to be adjusted based on the constant temperature and humidity collaborative adjustment instruction.

9. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored on the memory and executable on the processor. The processor executes the computer program to implement the constant temperature and humidity adjustment method for multi-region collaborative control according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium. When the computer program is executed by the processor, it implements the constant temperature and humidity adjustment method for multi-region collaborative control according to any one of claims 1-7.

Images