Agricultural park energy-saving management and control system and method based on cloud-side cooperation

Abstract

The invention discloses an agricultural park energy-saving management and control system and method based on cloud-side cooperation, and relates to the technical field of energy control. The side end controller obtains the data collected by the sensor, analyzes and processes the data, stores the data in a side end database, and uploads the data to the cloud server; obtaining an edge management and control model issued by the cloud server, and controlling the equipment based on the edge management and control model; and the side end management and control model determines an optimal scheduling scheme by adopting a particle swarm optimization algorithm by taking the lowest sub-region operation cost as an optimal scheduling objective function. According to the method, global optimization and local real-time control are combined through a cloud edge layered architecture, the data volume centrally processed by a cloud computing layer is remarkably reduced, and communication delay and distortion risks are reduced.

Description

Technical Field

[0001] This invention relates to the field of energy control technology, and in particular to an energy-saving management and control system and method for modern agricultural parks based on cloud-edge collaboration. Background Technology

[0002] Modern agricultural parks are a special form of agricultural production that applies modern technologies and management models, integrating various aspects such as planting, breeding, processing, and logistics. In recent years, the country has vigorously promoted the construction of modern agricultural parks. However, during the national energy structure transformation process, rural energy still suffers from problems such as low energy utilization efficiency and serious pollution. How to utilize collaborative optimization technology to achieve scheduling and optimized control among various nodes in the rural energy system, and efficiently realize task allocation, resource scheduling, and load balancing, urgently needs to be addressed. Summary of the Invention

[0003] The purpose of this invention is to provide a smart energy-using cloud-edge collaborative control system and control method for facility agriculture industrial parks, so as to realize intelligent control of the load of facility agricultural industrial parks in response to optimized control strategies and improve the renewable energy consumption rate.

[0004] Firstly, a cloud-edge collaborative energy-saving management and control system for agricultural parks is provided, comprising: a cloud server; and multiple smart greenhouses corresponding to various sub-regions of the agricultural park; each smart greenhouse includes:

[0005] Equipment, including energy storage devices, distributed power sources, and loads;

[0006] Sensors are used to collect data on equipment operating status and environmental parameters;

[0007] The edge controller is configured to: (1) acquire the data collected by the sensor, analyze and process it, store it in the edge database, and upload it to the cloud server; (2) acquire the edge control model issued by the cloud server, and control the equipment based on the edge control model; wherein the edge control model takes the minimum sub-region operating cost as the optimization scheduling objective function and uses the particle swarm optimization algorithm to determine the optimal scheduling scheme; wherein the operating cost includes the energy purchase cost, equipment operation and maintenance cost, wind and solar curtailment cost and flexible load compensation cost represented by the equipment model; wherein the optimal scheduling scheme determined by the particle swarm optimization algorithm satisfies power balance constraints, equipment output constraints and chemical cell constraints.

[0008] Secondly, a cloud-edge collaborative energy-saving management method for agricultural parks is provided. The agricultural park includes intelligent greenhouses set up for each sub-region. Each intelligent greenhouse includes: equipment, including energy storage devices, distributed power supplies, and loads; sensors for collecting equipment operating status and environmental parameter data; and an edge controller. The method includes the following steps executed by the edge controller:

[0009] The data collected by the sensor is acquired, analyzed, processed, stored in the edge database, and uploaded to the cloud server.

[0010] Obtain the edge control model issued by the cloud server, and control the device based on the edge control model;

[0011] The edge control model uses the lowest sub-region operating cost as the optimization scheduling objective function and employs a particle swarm optimization algorithm to determine the optimal scheduling scheme.

[0012] The operating costs include energy purchase costs, equipment operation and maintenance costs, wind and solar curtailment costs, and flexible load compensation costs, represented by equipment models; and the optimal scheduling scheme determined by the particle swarm optimization algorithm satisfies power balance constraints, equipment output constraints, and chemical cell constraints.

[0013] This invention combines global optimization with local real-time control through a cloud-edge layered architecture, which significantly reduces the amount of data centrally processed by the cloud computing layer and reduces communication latency and distortion risks.

[0014] Furthermore, by employing an improved particle swarm optimization algorithm and using dynamic constraint correction and model iteration mechanisms in the cloud, the edge control model is continuously optimized to solve the problem of collaborative optimization of multiple energy devices, taking into account both economic efficiency and the demand for renewable energy consumption, achieving efficient operation of multi-energy complementarity of cooling / heating / electricity, and improving energy utilization efficiency and system reliability.

[0015] In addition, the cloud-based model dynamic update mechanism supports unified management of heterogeneous equipment in multiple greenhouses, and the secure interaction scheme protects data privacy through encrypted transmission and partition protection, adapting to the complex network environment of agricultural parks. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the cloud-edge collaborative energy-saving management and control system architecture for agricultural parks according to an embodiment of the present invention;

[0017] Figure 2 An energy management architecture for a greenhouse multi-energy system based on edge computing;

[0018] Figure 3 This is a schematic diagram of the secure interaction architecture of the agricultural park energy cloud-edge management platform according to an embodiment of the present invention;

[0019] Figure 4 The values ​​represent the predicted values ​​of renewable energy and load for a typical summer day in a greenhouse. Figure 5 The results are optimized for edge computing. Detailed Implementation

[0020] The technical solution of the present invention will now be described in detail with reference to the accompanying drawings.

[0021] Based on factors such as the geographical location of distributed power sources and the load areas of facility agriculture, the agricultural park is divided into multiple sub-regions, namely the various greenhouses within the agricultural park. Each sub-region is equipped with an edge control node, which monitors and controls the energy status within the smart greenhouse through edge controllers. The edge layer is located within the greenhouse, enabling local control of smart energy. The results processed and analyzed at the edge layer are uploaded to the agricultural park. The cloud computing layer is located within the agricultural park, aggregating data from each edge layer, and then uploading the processed and analyzed results to the cloud platform.

[0022] Figure 1 This is a schematic diagram of an energy-saving management and control system architecture for agricultural parks based on cloud-edge collaboration, according to an embodiment of the present invention. Figure 1 As shown, it includes the cloud computing layer, communication layer, edge layer, and device layer.

[0023] The cloud computing layer, located within the agricultural park, performs global data analysis and optimization. It receives data uploaded from various edge layers, such as agricultural park data processed by edge nodes. By analyzing the data uploaded from each edge layer, the cloud computing layer optimizes energy management and control strategies.

[0024] In addition, the cloud computing layer generates global control signals based on the data uploaded by the edge nodes and the data analysis results, and sends them out through the edge controller to control the specific operations of each edge node.

[0025] In addition, the cloud computing layer can update and optimize various algorithm models in the agricultural park and distribute the updated algorithm models to each edge node to ensure that the edge nodes use the latest and best algorithm models for data processing and equipment control.

[0026] The communication layer uses network communication technology to connect the edge layer and the cloud computing layer, enabling efficient information exchange between the edge layer and the cloud computing layer, and ensuring the security and stability of data transmission.

[0027] The edge layer is configured in each sub-region of the agricultural park, namely each smart greenhouse. It is used for preprocessing edge data, such as aggregation and fusion, and for analyzing and processing edge data using algorithmic models (e.g., data processing models). Edge control nodes upload the processing and analysis results to the cloud computing layer. Furthermore, the edge layer processes data from each sub-region and generates corresponding control signals or updated data processing models based on the processing results. Simultaneously, the edge layer processes and analyzes the collected data locally within the agricultural park and sends control signals to each device, enabling the smart greenhouse energy equipment to achieve optimized energy control by executing appropriate control strategies.

[0028] The equipment layer comprises distributed power sources, loads, and energy storage devices configured within the smart greenhouse. Typical loads in a smart greenhouse include cold storage refrigeration units, plasma nitrogen fixation and water treatment loads, physical pest control loads, infrared heating loads, fertigation loads, and other agricultural power loads. Upon receiving control signals from the edge layer, the equipment layer executes corresponding control strategies. For example, it adjusts energy production, load consumption, or the operating status of energy storage devices. The equipment layer also includes sensors to collect data on equipment operating status and environmental parameters, and uploads this data to the edge layer for processing.

[0029] Traditional centralized computing models require concentrating large amounts of data in one or a few physical locations, with a central server providing computing and storage services. This approach suffers from high network transmission latency, making it difficult to meet the real-time multi-energy control requirements of smart greenhouses. For example, energy control in breeding greenhouses within agricultural parks needs rapid adjustments based on real-time monitoring data; however, high-latency data transmission leads to untimely control responses, directly impacting the crop's growth environment. Furthermore, traditional centralized computing models struggle to achieve efficient and comprehensive utilization of different energy sources. Smart greenhouses have diverse energy demands, such as electricity for lighting and equipment operation, heat for greenhouse heating, and cold energy for cold storage. These energy demands are unevenly distributed in time and space, requiring comprehensive consideration of the supply and demand of various energy sources for coordination and optimized scheduling. Therefore, this invention employs an edge computing-based multi-energy edge control method for smart greenhouses to meet the real-time business and energy management needs of smart greenhouses in agricultural parks.

[0030] like Figure 2 As shown, sensors collect information from various devices in the smart greenhouse and upload real-time environmental, load, and device information to the edge database. The edge database primarily uses wired and wireless networks for communication, employing interface protocols and interactive security technologies to achieve effective information acquisition, collaboration, transmission, and aggregation. The edge control model constructs the transmitted data into a fully operational and comprehensive digital carrier of the power grid system, enabling edge control of all greenhouse devices.

[0031] According to an embodiment of the present invention, in the edge layer scheduling mode, each smart greenhouse independently performs edge computing according to each sub-region under the optimization strategy given by the scheduling center, and performs optimization calculations on the terminal equipment in the smart greenhouse to obtain the optimal operation scheme of each energy device.

[0032] For example, the constructed edge autonomous system sub-region equipment model includes models of gas-fired boilers, waste heat boilers, electric refrigeration, absorption chillers, energy storage, photovoltaics, and wind turbines. To ensure the flexible operation of the edge autonomous sub-region system, energy storage devices can be configured to provide backup energy or absorb excess energy when energy supply and demand are unbalanced or energy fluctuations are significant, thereby maintaining system stability. Through the rational allocation and management of energy, the edge autonomous system sub-region can more effectively absorb distributed new energy sources while ensuring economic efficiency, achieving energy-saving management and control within the park.

[0033] The following provides exemplary representations of various equipment models, taking energy conversion equipment models and energy storage models as examples:

[0034] 1. Energy conversion equipment model

[0035] In intelligent greenhouses, energy conversion equipment mainly includes boilers as heating equipment and refrigeration units as cooling equipment, such as gas boilers, absorption chillers, electric chillers, and waste heat boilers.

[0036] (1) Gas-fired boiler

[0037] Gas-fired boilers are devices that convert gas energy into heat energy in the edge control sub-regions of agricultural parks. Their model can be represented as:

[0038] P GB (t)=P GB-in (t)η GB (1)

[0039] In the formula, P GB (t) and P GB-in (t) represents the heat output power of the gas-fired boiler and the natural gas input power at time t, respectively, η GB The thermal efficiency of a gas-fired boiler.

[0040] (2) Absorption chiller

[0041] Absorption chillers are the main cooling equipment in the periphery control sub-areas of agricultural parks, converting input heat into output cooling. Their model can be represented as:

[0042] P AC (t)=P AC-in (t)α AC (2)

[0043] In the formula, αAC P represents the cold output efficiency ratio of an absorption chiller. AC (t), P AC-in (t) represents the cold output power and heat input power of the absorption chiller at time t, respectively.

[0044] (3) Electric Refrigeration Machine (EC)

[0045] Electric chillers are also important energy conversion devices in the edge control sub-regions of agricultural parks, and their model can be represented as:

[0046] P EC-C (t)=P EC (t)α EC (3)

[0047] In the formula, P EC-C (t) represents the cold energy output of EC during time period t, P EC (t) represents the electrical energy consumption of EC during time period t, α EC This is the energy efficiency coefficient of EC.

[0048] (4) Waste heat boiler

[0049] The waste heat recovery boiler collects and reheats the waste heat emitted by the gas-fired boiler in the controlled sub-area at the edge of the greenhouse. The energy flow coupling conversion model is as follows:

[0050] P HR (t)=P GB (t)α GB ·η HR (4)

[0051] In the formula, P HR (t) represents the heating power of the waste heat boiler, η HR This refers to the heating efficiency of waste heat boilers.

[0052] (2) Energy storage model

[0053]

[0054] Where, E ESS (t) represents the energy storage capacity during time period t, P store (t), P release (t), η store η release These represent the charging and discharging power and efficiency during time period t, respectively.

[0055] The objective function for optimizing the scheduling of the edge control system is to minimize the operating cost of the sub-region of the edge autonomous system. The operating cost mainly includes the operation and maintenance costs of each unit, the penalty cost for unused renewable energy, energy purchase costs, and flexible load compensation costs. The model is as follows:

[0056] CIES (t)=C GE (t)+C CUT (t)+C YW (t)+C FH (t) (6)

[0057] In the formula, C IES (t) represents the operating cost during time period t; C GE (t) represents the energy purchase cost during time period t; C YW (t) represents the operation and maintenance cost of each device during time period t; C CUT (t) represents the cost of curtailing wind and solar power during time period t; C FH (t) represents the cost of flexible load compensation.

[0058] in,

[0059] C GE (t)=P buy-e (t)M e (t)+P GB (t)M g (7)

[0060] C CUT (t)=M cut P loss (t) (8)

[0061] C YW (t)=K v P v (t) (9)

[0062] C FH (t)=F shift +F cut (10)

[0063] In the formula, P buy-e (t) represents the system's electricity purchase at time t; M e (t) represents the electricity purchase price during time period t; M g For natural gas prices; M cut The cost per unit of wind and solar curtailment; P loss (t) represents the amount of wind and solar power curtailed during time period t; K v P represents the unit operation and maintenance cost coefficient, where v is the unit type; v (t) represents the output of unit v at time t; F shift 、F cut These are the compensation costs for transferable cooling, heating, and electrical loads and the costs for reducing cooling, heating, and electrical loads when they participate in system optimization.

[0064] The constraints include:

[0065] (1) Power balance constraint

[0066]

[0067] In the formula, L ele (t), L cool (t), L heat (t) represent the electrical, cooling, and heating loads of the integrated energy system, respectively; P WT (t), P PV (t) represents the output of the wind turbine and the photovoltaic system, respectively; P GRID (t) represents the interaction power between the integrated energy system and the power grid (greater than 0 represents electricity purchase, less than 0 represents electricity sale); P BAT (t) represents the charge and discharge power of the chemical battery (greater than 0 indicates charging, less than 0 indicates discharging).

[0068] (2) Equipment output constraints

[0069] P i-min ≤P i ≤P i-max (12)

[0070] In the formula, P i-min P i-max These represent the lower and upper limits of the output of the i-devices, including energy conversion equipment and energy storage.

[0071] (3) Chemical cell confinement

[0072] E ESS-min ≤E ESS ≤E ESS-max (13)

[0073] Where, E ESS Energy storage capacity; E ESS-min E ESS-max These are the lower and upper limits for storing electrical energy, respectively.

[0074] As an example, when solving the edge-side control optimization problem, an improved Particle Swarm Optimization (PSO) algorithm is used to solve the optimization scheduling model of the multi-energy greenhouse edge control system in the agricultural park. The main steps of the improved PSO algorithm in solving the greenhouse edge control system include:

[0075] (1) Initialize the basic parameters of the particles;

[0076] Set the maximum number of algorithm iterations, particle dimension, particle swarm size, self-learning factor, and inertia weight, and randomly set the velocity and position of each particle.

[0077] (2) Obtain parameters of energy conversion equipment, renewable energy output, and greenhouse load parameters of the greenhouse edge control system;

[0078] (3) Calculate the fitness value of each particle, i.e. the total running cost, and update the historical best position and global best position of the particles;

[0079] (4) Update the velocity and position of all particles;

[0080] (5) Determine and adjust the solution of the particle to satisfy the constraints such as power balance, equipment output and energy storage;

[0081] (6) Iterate and update until the maximum number of iterations is reached or the global best position no longer changes significantly. Output the global best position and its corresponding fitness value as the optimal scheduling scheme.

[0082] Smart greenhouses generate a large amount of data every day, such as environmental data (e.g., temperature, humidity, light intensity), energy equipment status data, and energy consumption data. Edge computing technology can transfer data processing and computing functions from traditional centralized data centers to devices at the network edge. On the one hand, data can be processed locally at the greenhouse, but on the other hand, due to the limited storage capacity of the edge database, it can only store short-term data.

[0083] Furthermore, agricultural parks often consist of various intelligent greenhouses or agricultural workshops. For example, a modern potato agricultural park may include a new variety breeding workshop, a spray cultivation workshop, a tissue culture workshop, multi-span greenhouses, and digital intelligent greenhouses. The energy equipment and data types of each greenhouse may differ, and different sensors and control systems may be used, posing challenges to the rational energy use and overall operation management of the agricultural park. Therefore, the general edge control model constructed above needs to be iterated and modified according to the specific characteristics of each greenhouse.

[0084] To address the aforementioned issues, cloud-edge collaborative operation is needed to optimize the effectiveness of edge-only control. Specifically, in actual production, agricultural parks consider energy data from various greenhouses, aggregating data from multiple greenhouses. Simultaneously, cloud servers deployed within the agricultural park are used for preliminary analysis and optimization decisions. By leveraging a large amount of historical operational data stored within the park, and through cloud computing, targeted improved edge-edge control models are established for each greenhouse based on a general edge-edge control model. These models are then distributed to edge controllers, enabling continuous iterative optimization of control strategies. This achieves efficient management and optimization of the agricultural park's comprehensive energy system, improving energy utilization and production efficiency.

[0085] In terms of cloud-edge interaction, since the smart terminals in agricultural parks lack an adaptive connection system with cloud servers, the terminal devices may use different communication technologies and protocols, such as Wi-Fi, LoRa, MQTT, etc. To ensure that data can be transmitted and processed smoothly, a cloud-edge interaction solution compatible with various communication protocols needs to be designed.

[0086] Meanwhile, agricultural parks involve a large amount of agricultural production data, including crop growth information and environmental monitoring data, requiring the cloud-edge interaction solution to ensure the security and privacy protection of data transmission.

[0087] In addition, agricultural parks are often located in remote areas where network bandwidth and latency may be problematic, affecting the efficiency and timeliness of data interaction. Therefore, it is necessary to design a suitable cloud-edge interaction solution to address the challenges of the network environment.

[0088] Figure 3 This is a schematic diagram of the secure interaction architecture of the agricultural park energy cloud-edge management platform according to an embodiment of the present invention. Figure 3 As shown, the agricultural park energy cloud-edge management platform system is divided into internal and external network areas. Different security protection requirements are defined for different areas within the internal and external networks. To ensure secure interaction of the modern agricultural park energy cloud-edge management platform, security authentication platforms and firewalls are deployed for different security zones. Firewalls are deployed to isolate the internal and external networks. Data collected by intelligent terminal devices in the agricultural park is encrypted and connected to a secure access platform, and then authenticated before being connected to the power grid company's IoT management platform.

[0089] The power grid company's IoT management platform connects multiple edge control devices, enabling intelligent regulation of edge acquisition and control devices. These devices collect environmental data, collect electricity, cooling, gas, and heat metering data from users in agricultural parks, control and optimize the load of facility agriculture, and can also be used for distributed energy management and power quality analysis in agricultural parks.

[0090] Edge acquisition and control devices, in turn, provide execution feedback to edge management devices. Data generated within the agricultural park is captured, unified under the power grid company's IoT management platform, and then transmitted to the modern agricultural park's energy cloud edge management platform. Applications such as optimized control, organic fertilizer processing, agricultural breeding, fault alarms, agricultural tourism, and mushroom production are all deployed on the agricultural park cloud platform. The modern agricultural park's energy cloud edge management platform can access its functions via intranet terminals. Intranet server applications are deployed on the power grid company's intranet, utilizing cloud resources. Simultaneously, an external network entry point is deployed on an external service website for use by on-site personnel in the agricultural park.

[0091] In summary, this invention proposes a cloud-edge collaborative energy-saving management architecture for modern agricultural parks. The cloud computing layer is deployed within the agricultural park for global data aggregation, strategy optimization, and model iteration updates. The edge layer is deployed in each intelligent greenhouse sub-area, enabling real-time monitoring and optimized control of local energy equipment through edge controllers. The equipment layer includes distributed power supplies, cooling / heating / electrical loads, and energy storage devices, executing control commands issued by the edge layer. The communication layer uses a secure transmission protocol to connect the cloud, edge, and device layers, ensuring low latency and reliability of data interaction.

[0092] Furthermore, this invention proposes an edge-autonomous optimization scheduling model based on an improved particle swarm optimization algorithm. It constructs an energy efficiency coupling model and dynamic constraints for multiple energy devices (gas boilers, waste heat boilers, electric chillers, energy storage, etc.), takes the minimization of operating costs as the objective function, comprehensively considers energy purchase costs, new energy curtailment penalties, and flexible load compensation costs, and dynamically corrects the constraints through an improved particle swarm optimization algorithm, ultimately achieving coordinated optimization of power balance, equipment output, and energy storage capacity.

[0093] In addition, this invention designs a dynamic update mechanism for the cloud-edge collaborative model. The cloud generates a targeted edge control model based on the aggregation and analysis of historical data from multiple greenhouses and distributes it to the edge layer. The edge layer feeds back local operating data to the cloud in real time, realizing online calibration of model parameters and iterative optimization of strategies, thus solving the problem of collaborative control of heterogeneous equipment in multiple greenhouses.

[0094] Meanwhile, the agricultural park energy cloud-edge security interaction solution proposed in this invention adopts an internal and external network partitioning architecture, realizes encrypted data transmission through firewall isolation and security authentication platform, is compatible with multi-protocol terminal access, supports unified aggregation and management of heterogeneous device data, and ensures data privacy and system stability.

[0095] Example:

[0096] We selected actual operating data of the intelligent greenhouse integrated energy system in an agricultural park for analysis, and built an integrated energy physical model of the park on the CloudPSS platform to simulate the equipment operating status and energy flow in the actual environment in order to verify the edge control strategy.

[0097] This intelligent greenhouse includes cold storage refrigeration units, plasma nitrogen fixation and water treatment loads, physical pest control loads, infrared heating loads, integrated water and fertilizer loads, and other agricultural power loads. Typical summer greenhouse renewable energy and load forecasts are as follows: Figure 4 As shown, the intelligent greenhouse system of the integrated energy system in a modern agricultural park consists of park terminal equipment, energy equipment, edge database, and various sensors.

[0098] A physical simulation model of an integrated energy system was built based on the CloudPSS simulation platform. An improved particle swarm optimization algorithm was implemented by calling the CloudPSS simulation platform through Python programming. The CloudPSS simulation platform can not only realize the operation optimization and scheduling control of integrated energy systems, but also realize efficient and fast energy flow simulation calculations for edge control systems.

[0099] Analysis of edge computing optimization results, such as Figure 5 As shown, the energy generated by the power supply equipment and the electricity purchased from the external grid are represented as positive values, while the energy consumed by the equipment is represented as negative values. Similarly, the energy released by the energy storage equipment into the system is represented as positive values, while the stored energy is represented as negative values.

[0100] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein, and such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An energy-saving management and control system for agricultural parks based on cloud-edge collaboration, characterized in that, include: Cloud server; And multiple smart greenhouses corresponding to each sub-area of ​​the agricultural park; Each smart greenhouse includes: Equipment, including energy storage devices, distributed power sources, and loads; Sensors are used to collect data on equipment operating status and environmental parameters; The edge controller is configured to: (1) acquire the data collected by the sensor, analyze and process it, store it in the edge database, and upload it to the cloud server; (2) acquire the edge control model issued by the cloud server, and control the equipment based on the edge control model; wherein the edge control model takes the minimum sub-region operating cost as the optimization scheduling objective function and uses the particle swarm optimization algorithm to determine the optimal scheduling scheme; wherein the operating cost includes the energy purchase cost, equipment operation and maintenance cost, wind and solar curtailment cost and flexible load compensation cost represented by the equipment model; wherein the optimal scheduling scheme determined by the particle swarm optimization algorithm satisfies power balance constraints, equipment output constraints and chemical cell constraints.

2. The system according to claim 1, wherein: Operating cost C during period t IES (t) is: C IES (t)=C GE (t)+C CUT (t)+C YW (t)+C FH (t) Among them, C GE (t) represents the energy purchase cost during time period t, C GE (t)=P buy-e (t)M e (t)+P GB (t)M g C YW (t) represents the operation and maintenance cost of each device during time period t, C YW (t)=K v P v (t); C CUT (t) represents the cost of wind and solar power curtailment during time period t, C CUT (t)=M cut P loss (t); C FH (t) represents the flexible load compensation cost, C FH (t)=F shift +F cut ; In the formula, P buy-e (t) represents the electricity purchased at time t; M e (t) represents the electricity purchase price during time period t; P GB For the gas-fired boiler model, M represents the heat output power of the gas-fired boiler at time t; g For natural gas prices; M cut The cost per unit of wind and solar curtailment; P loss (t) represents the amount of wind and solar power curtailed during time period t; K v P represents the unit operation and maintenance cost coefficient, where v is the unit type; v (t) represents the output of unit v at time t; F shift 、F cut These are the compensation costs for transferable cooling, heating, and electrical loads and the costs for reducing cooling, heating, and electrical loads when they are involved in optimization.

3. The system according to claim 2, characterized in that, The power balance constraint is: In the formula, L ele (t), L cool (t), L heat (t) represents the electrical, cooling, and heating loads of the system during time period t; P WT (t), P PV (t) represents the output of the wind turbine and the photovoltaic system during time period t, respectively; P GRID (t) represents the interaction power between the system and the power grid during time period t; P BAT (t) represents the charge / discharge power of the chemical battery during time period t; P EC (t) represents the electrical energy consumption of the electric chiller during time period t; P EC-C (t) represents the electric chiller model, indicating the cooling energy output by the electric chiller during time period t; P AC (t) represents the absorption chiller model, indicating the cold output power of the absorption chiller at time t.

4. The system according to claim 3, characterized in that, The equipment output constraint is that the equipment output is within an upper and lower limit range, and the chemical battery constraint is that the stored electrical energy is within an upper and lower limit range.

5. The system according to claim 1, wherein: The edge control model is obtained by optimizing the general edge control model for each smart greenhouse based on the data uploaded by the edge controller by the cloud server.

6. A cloud-edge collaborative energy-saving management method for agricultural parks, wherein, The agricultural park includes intelligent greenhouses set up for each sub-area. Each intelligent greenhouse includes: equipment, including energy storage devices, distributed power sources, and loads; sensors for collecting equipment operating status and environmental parameter data; and an edge controller. The method is characterized by the following steps executed by the edge controller: The data collected by the sensor is acquired, analyzed, processed, stored in the edge database, and uploaded to the cloud server. Obtain the edge control model issued by the cloud server, and control the device based on the edge control model; The edge control model uses the lowest sub-region operating cost as the optimization scheduling objective function and employs a particle swarm optimization algorithm to determine the optimal scheduling scheme. The operating costs include energy purchase costs, equipment operation and maintenance costs, wind and solar curtailment costs, and flexible load compensation costs, represented by equipment models; and the optimal scheduling scheme determined by the particle swarm optimization algorithm satisfies power balance constraints, equipment output constraints, and chemical cell constraints.

7. The method according to claim 6, characterized in that, Operating cost C during period t IES (t) is: C IES (t)=C GE (t)+C CUT (t)+C YW (t)+C FH (t) Among them, C GE (t) represents the energy purchase cost during time period t, C GE (t)=P buy-e (t)M e (t)+P GB (t)M g C YW (t) represents the operation and maintenance cost of each device during time period t, C YW (t)=K v P v (t); C CUT (t) represents the cost of wind and solar power curtailment during time period t, C CUT (t)=M cut P loss (t); C FH (t) represents the flexible load compensation cost, C FH (t)=F shift +F cut ; In the formula, P buy-e (t) represents the electricity purchased at time t; M e (t) represents the electricity purchase price during time period t; P GB For the gas-fired boiler model, M represents the heat output power of the gas-fired boiler at time t; g For natural gas prices; M cut The cost per unit of wind and solar curtailment; P loss (t) represents the amount of wind and solar power curtailed during time period t; K v P represents the unit operation and maintenance cost coefficient, where v is the unit type; v (t) represents the output of unit v at time t; F shift 、F cut These are the compensation costs for transferable cooling, heating, and electrical loads and the costs for reducing cooling, heating, and electrical loads when they are involved in optimization.

8. The method according to claim 7, characterized in that, The power balance constraint is: In the formula, L ele (t), L cool (t), L heat (t) represents the electrical, cooling, and heating loads of the system during time period t; P WT (t), P PV (t) represents the output of the wind turbine and the photovoltaic system during time period t, respectively; P GRID (t) represents the interaction power between the system and the power grid during time period t; P BAT (t) represents the charge / discharge power of the chemical battery during time period t; P EC (t) represents the electrical energy consumption of the electric chiller during time period t; P EC-C (t) represents the electric chiller model, indicating the cooling energy output by the electric chiller during time period t; P AC (t) represents the absorption chiller model, indicating the cold output power of the absorption chiller at time t.

9. The method according to claim 8, characterized in that, The equipment output constraint is that the equipment output is within an upper and lower limit range, and the chemical battery constraint is that the stored electrical energy is within an upper and lower limit range.

10. The method according to claim 6, characterized in that, The edge control model is obtained by optimizing the general edge control model for each smart greenhouse based on the data uploaded by the edge controller by the cloud server.

Images