Energy-saving refrigeration storage system

Abstract

The invention discloses an energy-saving refrigeration storage system. The system comprises an environment sensing module, a solar power generation module, an air energy heat pump module, a cold storage module, an energy efficiency modeling and intelligent regulation and control module and a remote monitoring and feedback module. The system utilizes a multi-source fusion control mechanism to autonomously optimize a refrigeration strategy and an energy scheduling path based on an intelligent regulation and control strategy of a reinforcement learning algorithm according to information such as the solar irradiance, the environment temperature, the internal temperature of the refrigeration house and the cold load demand, and the comprehensive energy efficiency ratio and the operation economy are improved. The system is suitable for the scenes of agricultural product cold chain and low-temperature storage, and has the characteristics of remarkable energy-saving effect, high operation intelligence degree and strong adaptability.

Description

Technical Field

[0001] The present invention relates to an energy-saving refrigeration cold storage system based on an air energy heat pump, and in particular to a system that utilizes solar energy and air heat sources for refrigeration and energy storage. Background Art

[0002] As the global energy crisis and pressure to control carbon emissions intensify, cold storage, as an energy-intensive facility, faces an increasingly urgent need for energy-saving and clean transformation. Currently, cold storage systems widely use electric-driven compression refrigeration, which presents challenges such as single energy utilization, concentrated peak energy consumption, and rising operating costs. This is particularly true in large-scale cold chain scenarios, such as those for agricultural products, where low energy scheduling efficiency and large load fluctuations pose multiple challenges to the operational stability and economic efficiency of refrigeration systems.

[0003] To reduce reliance on traditional electric energy, some cold storage systems have incorporated solar energy or air-source heat pump technology to form a supplementary clean energy structure. However, most existing solutions only achieve simple energy aggregation and fail to achieve dynamic energy efficiency synergy based on factors such as sunlight fluctuations, temperature differences, and changes in cooling load. Furthermore, solar energy resources are highly time-varying and intermittent, and are not effectively integrated with energy storage mechanisms, resulting in low renewable energy utilization. Some systems have introduced cold storage technology, but widespread issues include extensive control over cold storage timing, mismatched storage and discharge, and an inability to effectively participate in peak load shifting and shaving.

[0004] In addition, most current solar heat pump systems still rely on static settings or manual experience judgments in their control strategies. They lack optimization mechanisms based on environmental information perception, adaptive learning, and intelligent scheduling capabilities, making it difficult to cope with the dynamic complexity of cold storage operations. This results in limited improvements in system energy efficiency and makes it difficult to balance economy and stability.

[0005] For example, CN116697639A in the prior art discloses a solar-air source heat pump coupled hot water, heating, cooling, and energy storage system and its control method, which has the following problems: the control logic relies on a single control quantity of the temperature sensor, and the control method mainly relies on two temperature sensors (T1 and T2). The perception parameter dimensions are limited, and it is impossible to accurately judge the changes in multiple heat sources / loads, which may lead to switching misjudgments or delays; the energy storage method is single and the scheduling mechanism is vague: although there is an energy storage mode, it only relies on a buffer water tank and does not involve other high-efficiency energy storage media or energy-saving mechanisms (such as PCM phase change materials, electrothermal coupled energy storage, etc.), and the overall system dynamic load adjustment capability is insufficient.

[0006] Therefore, it is urgent to propose an intelligent cold storage energy efficiency management system for the collaborative work of solar energy and air source heat pumps, which can realize multi-energy integration, autonomous optimization and regulation, intelligent cold storage and release, and stable and safe operation to meet the current application needs of green, efficient and intelligent development of cold storage. Summary of the Invention

[0007] The purpose of the present invention is to provide an energy-saving refrigeration cold storage system based on an air energy heat pump, and in particular to a system that utilizes solar energy and air heat sources for refrigeration and energy storage, which can achieve multi-energy integration, autonomous optimization and regulation, intelligent cold storage and release, and stable and safe operation. To achieve the above purpose, the present invention provides the following technical solutions:

[0008] An energy-saving refrigeration cold storage system, characterized by comprising:

[0009] Environmental sensing module, used to collect real-time environmental and operating parameters such as solar irradiance, ambient temperature, internal temperature of cold storage, cooling load demand, and photovoltaic power generation output;

[0010] Solar power generation module, used to convert solar radiation into AC electrical energy for use by the system;

[0011] Air energy heat pump module, used to provide cooling or heating energy according to control instructions;

[0012] The cold storage module is used to store cold energy during off-peak hours or when there is sufficient solar energy, and release the cold energy during high-load hours;

[0013] Energy efficiency modeling and intelligent control module, which includes:

[0014] Energy modeling submodule, used to establish a multi-source energy flow mathematical model including photovoltaic power generation, cold storage thermal characteristics, air energy heat pump efficiency, and the charging and discharging dynamics of cold storage units;

[0015] The intelligent scheduling submodule, based on the reinforcement learning algorithm, inputs environmental status information and outputs an optimized energy distribution control strategy to minimize the system energy consumption cost;

[0016] The remote monitoring and feedback module is used to access the cloud platform to achieve remote operation management, data visualization and energy efficiency analysis.

[0017] Optionally, the air-to-heat pump module mainly includes a compressor, an evaporator, a condenser, an expansion valve, a fan and water pump, and a controller.

[0018] Optionally, the cold storage module includes a cold storage tank, a phase-change cold storage medium, and a heat exchange unit.

[0019] Optionally, the energy modeling submodule adopts a multivariable dynamic modeling method to establish a cold storage temperature change equation, a cold storage energy change equation and a comprehensive power balance relationship for predicting system state transition.

[0020] Optionally, the state space set by the reinforcement learning algorithm is:

[0021] S t =[G(t),Tamb (t),T in (t),Q load (t),E cs (t),P pv (t),TimeOfDay]

[0022] Where: G(t) is the solar irradiance, T amb (t) is the ambient temperature, T in (t) is the internal temperature of the cold storage, Q load (t) is the cooling load forecast, E cs (t) is the current cold storage energy, P pv (t) is the photovoltaic power output, and TimeOfDay is the current time information.

[0023] Optionally, the action space set by the reinforcement learning algorithm includes the following three continuous control variables:

[0024] Air source heat pump operating power control variable u ashp , the range is [0,1];

[0025] Cold storage charging control variable u ch , the range is [0,1];

[0026] Cooling energy storage control variable u dis , the range is [0,1].

[0027] Optionally, the reward function of the reinforcement learning algorithm is the negative value of the total electricity cost of the system.

[0028] Optionally, the reinforcement learning algorithm of the intelligent scheduling submodule adopts deep deterministic policy gradient (DDPG), and its reinforcement learning process is specifically as follows:

[0029] Initialize the strategy network and value network;

[0030] Initialize the experience replay pool and target network;

[0031] For each task cycle:

[0032] a. The environment perception module gives the current state space S t ;

[0033] b. Policy network output action space A t ;

[0034] c. Calculate the next state S through the energy modeling submodule t+1 , get the reward function R t ;

[0035] d. Storage (S t ,At ,R t ,S t+1 );

[0036] e. Sample batches from the experience pool for gradient update;

[0037] Use experience replay and target network training strategies;

[0038] Deploy after training until the reward function converges.

[0039] Optionally, remotely set temperature set points and energy efficiency policy constraints; display current energy consumption structure, cold storage status and scheduling logs; upload data to the cloud platform, support historical data backtracking and energy efficiency evaluation report output.

[0040] A cold storage device adopts any of the above cold storage systems.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] 1. Significantly reduce energy costs. Reinforcement learning algorithms are used to schedule heat pumps, cold storage, and photovoltaic output in real time, prioritizing the use of low-price electricity and self-generated photovoltaic power, reducing peak-period electricity purchases and minimizing overall system electricity costs.

[0043] 2. Improve the self-consumption rate of solar energy. The environmental sensing module monitors irradiance and power generation in real time. Combined with intelligent scheduling, it automatically stores energy or directly drives loads during periods of sufficient sunlight, maximizing the utilization of photovoltaic power generation and avoiding wasteful solar power.

[0044] 3. Peak shaving and valley filling, load smoothing. The cold storage module stores cold energy during off-peak hours or when there is surplus photovoltaic power, and releases it during peak hours, effectively smoothing fluctuations in heat pump and grid loads, improving grid access stability and equipment lifespan.

[0045] 4. Efficient refrigeration and stable temperature control. Using a cold storage thermal model, the system meets cooling load requirements in real time and maintains temperature deviation to ensure product quality.

[0046] 5. Dynamic self-adaptation and multi-objective optimization. Based on deep reinforcement learning, the system can self-learn peak and valley electricity prices, weather, and load variations, achieving multi-objective optimization control of energy costs, carbon emissions, and equipment lifespan. The strategy is continuously optimized as the environment evolves.

[0047] 6. Modular design, easy to expand and integrate. Each functional module (environmental sensing, photovoltaic power generation, heat pump, cold storage, energy efficiency modeling, remote monitoring) is highly decoupled, supporting on-demand addition, subtraction, or upgrade, facilitating integration into cold storage, warehousing, and industrial refrigeration scenarios of varying sizes.

[0048] 7. Convenient remote monitoring and operation and maintenance. Cloud platform access enables real-time data visualization, remote parameter distribution, and fault warnings, greatly improving system operation and maintenance efficiency and management transparency. It supports multi-terminal access from mobile devices and industrial touch terminals.

[0049] 8. Reduce carbon emissions and achieve green and low-carbon development. Replacing traditional electric heating / cooling with solar energy and high-efficiency heat pumps achieves synergy between clean energy and energy storage, significantly reducing the carbon footprint of system operation and meeting energy conservation, emission reduction, and carbon neutrality requirements.

[0050] Through the above beneficial effects, the present invention can provide efficient, reliable and sustainable intelligent energy efficiency management solutions for the fields of cold chain, preservation, low-temperature processing, etc. at both technical and economic levels. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 Schematic diagram of the overall framework of the system in an embodiment of the present invention;

[0052] Figure 2 Schematic diagram of the visual interface of the system in an embodiment of the present invention. DETAILED DESCRIPTION

[0053] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0054] In the description of the present invention, it should be understood that the terms "upper", "lower", "front", "back", "left", "right", "top", "bottom", "inside", "outside", etc., indicating directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore should not be understood as limiting the present invention.

[0055] Reference Figure 1 The system framework diagram is shown in FIG. , and the embodiments of the present disclosure are described in detail below in conjunction with the accompanying drawings.

[0056] An energy-saving refrigeration cold storage system, characterized by comprising:

[0057] Environmental sensing module, used to collect real-time environmental and operating parameters such as solar irradiance, ambient temperature, internal temperature of cold storage, cooling load demand, and photovoltaic power generation output;

[0058] Solar power generation module, used to convert solar radiation into DC or AC electricity for system use;

[0059] Air energy heat pump module, used to provide cooling or heating energy according to control instructions;

[0060] The cold storage module is used to store cold energy during off-peak hours or when there is sufficient solar energy, and release the cold energy during high-load hours;

[0061] Energy efficiency modeling and intelligent control module, which includes:

[0062] Energy modeling submodule, used to establish a multi-source energy flow mathematical model including photovoltaic power generation, cold storage thermal characteristics, air energy heat pump efficiency, and the charging and discharging dynamics of cold storage units;

[0063] The intelligent scheduling submodule, based on the reinforcement learning algorithm, inputs environmental status information and outputs an optimized energy distribution control strategy to minimize the system energy consumption cost;

[0064] The remote monitoring and feedback module is used to access the cloud platform to achieve remote operation management, data visualization and energy efficiency analysis.

[0065] Among them, the environmental perception module is used to collect environmental and operating parameters such as solar irradiance, ambient temperature, internal temperature of cold storage, cooling load demand and photovoltaic power generation output in real time.

[0066] Specifically, the environmental perception module is the information input core of this system. Its main function is to perceive the changes in the external environment and the internal operating status of the cold storage in real time through multi-source sensors and data interfaces, providing accurate and timely data support for subsequent energy efficiency modeling and intelligent scheduling.

[0067] Its perception dimensions include but are not limited to the following five key parameters:

[0068] 1. Solar irradiance G(t), which is used to reflect the solar radiation intensity received per unit area at the current moment (unit: W / m 2 ) is an important basis for assessing the potential of photovoltaic power generation. This parameter is acquired in real time through light sensors or the photovoltaic inverter's built-in monitoring module, supporting dynamic prediction of photovoltaic power generation capacity.

[0069] 2. Ambient temperature T a (t) Ambient temperature is a key factor affecting the heat exchange efficiency of air-source heat pumps, especially in winter and summer when temperature differences are large, significantly impacting the system's energy efficiency. This parameter is collected using standard temperature and humidity sensors, supporting local measurement or access to meteorological platform data.

[0070] 3. Cold storage internal temperature T i (t) is used to monitor the current temperature of the cold storage and ensure the stability of the cold storage environment. Temperature data collected from temperature collection points throughout the cold storage space can be used to determine the cooling load demand and the effectiveness of the heat pump scheduling strategy.

[0071] 4. Cooling load demand Q(t). Cooling load refers to the cooling capacity (in kW) required to maintain the set temperature at the current moment. It is determined by factors such as the temperature difference between the inside and outside of the cold storage, the frequency of material inflow and outflow, and the frequency of door opening and closing. This can be estimated in real time using a heat balance model or in the short term using a load forecasting model trained on historical operating conditions.

[0072] 5. Photovoltaic power output P_pv(t) is the actual power output of the PV array at the current moment (unit: kW), reflecting the clean energy supply capacity. This data can be collected in real time through the communication interface provided by the PV inverter and can also be connected to the energy management system (EMS) for data integration.

[0073] The environmental perception module transmits collected data to the central controller via wireless or wired means and supports the following features:

[0074] Edge computing capabilities: It has preliminary data filtering, anomaly detection and preprocessing capabilities to improve communication efficiency; data synchronization and caching mechanism: supports slow transmission during network disconnection and redundant backup to ensure system stability; cloud visualization access: perception data can be uploaded to the cloud platform for remote monitoring and algorithm training.

[0075] Among them, the solar power generation module is used to convert solar radiation into AC electrical energy for system use.

[0076] Specifically, the solar power generation module is the core of the green energy source of this system. Its main function is to convert the collected solar radiation energy into DC or AC electricity that can be used by the system, which is used to drive air-source heat pumps, cold storage units or directly supply cold storage loads, thereby effectively reducing the cost of purchasing electricity from the power grid and improving energy self-sufficiency.

[0077] This module mainly includes the following key sub-units:

[0078] Photovoltaic module arrays. Photovoltaic modules are the front-end equipment for solar power generation. They typically use crystalline silicon (monocrystalline or polycrystalline) solar panels, installed in an array configuration on open areas such as cold storage roofs, floors, or facades. The output voltage is DC, and the voltage level is related to the series and parallel connection of the modules.

[0079] The photovoltaic inverter converts the DC power output from the photovoltaic modules into standard AC power (e.g., AC220V / 380V) to meet the power requirements of other system modules. Its functions include: MPPT (maximum power point tracking) control to improve power generation efficiency; power quality management (anti-islanding protection, harmonic filtering, etc.); a communication interface with the energy efficiency management controller to facilitate scheduling and management; and grid-connected and off-grid switching capabilities (depending on the application scenario).

[0080] Photovoltaic output monitoring and data acquisition interface. This part is used to obtain key operating parameters of the photovoltaic system in real time, including: current output power P_pv(t), voltage, current, cumulative power generation value, and real-time efficiency. The monitoring data can be directly used as input variables for the system's energy efficiency analysis and scheduling algorithm.

[0081] Power output paths and supply logic. Based on the system scheduling strategy, PV power output can achieve the following paths: prioritize supplying heat pump operating loads; charging cold storage units; optionally feeding excess power back to the grid; and coordinating with grid power sources to achieve hybrid power supply control.

[0082] Optionally, the air-to-heat pump module mainly includes a compressor, an evaporator, a condenser, an expansion valve, a fan and water pump, and a controller.

[0083] Specifically, the air-to-air heat pump module is the core cold and heat source device of this system. It is used to provide the required cooling or heating energy under different operating conditions according to the power instructions issued by the controller to meet the temperature control and cold storage requirements of the cold storage. This module mainly includes the following key subunits and functions:

[0084] 1. Main equipment of refrigeration cycle

[0085] The compressor uses a variable frequency scroll or screw compressor, combined with variable frequency drive (VFD) technology, which can accurately adjust the cooling capacity output.

[0086] The evaporator (evaporative heat exchanger) absorbs heat from the ambient air, causing the refrigerant to evaporate into a gaseous state. It is the heat absorption end of the heat pump.

[0087] The condenser (condensing heat exchanger) condenses the high-pressure and high-temperature refrigerant gas and releases heat to provide cooling for the cold storage or cold storage tank.

[0088] The expansion valve, either an electronic expansion valve or a thermal expansion valve, controls the refrigerant throttling amount to ensure that the evaporator always operates under the optimal superheat condition.

[0089] Fan and water pump. The fan is mainly used in the evaporator to introduce external air into the heat pump system for heat exchange with the refrigerant. The water pump is used in the condenser to transport cold / hot water to the terminal condenser and is responsible for the charging and discharging of the refrigerant in the cold storage system.

[0090] 2. Control and sensor unit

[0091] Electrical controller (PLC / MCU), receives P output from central intelligent dispatch module ashp (t) or normalized control quantity is converted into the drive signal of the inverter and expansion valve to achieve refined power regulation.

[0092] The sensor unit includes:

[0093] Temperature and pressure sensors: Evaporator inlet and outlet temperatures, condenser inlet and outlet temperatures, refrigerant pressure (high-pressure side and low-pressure side), and cold storage return air temperature. These sensor data are used to monitor the heat pump's operating status in real time and provide feedback to the controller.

[0094] Flow meters are used to measure the chilled water (or air duct) flow rate in order to accurately calculate the actual heat transfer.

[0095] Optionally, the cold storage module includes a cold storage tank, a phase-change cold storage medium, and a heat exchange unit.

[0096] Specifically, the cold storage module is the key energy storage unit of this system. It stores cooling capacity when electricity prices are low or photovoltaic output is sufficient, and releases cooling energy when the cooling load is peak or there is insufficient sunlight, thereby achieving "peak shaving and valley filling" of the system and optimizing energy utilization.

[0097] 1. Structural composition

[0098] The cold storage tank (cold storage medium) uses water / ethylene glycol solution, ice storage or phase change material (PCM) and other media. It is designed as a multi-layer heat exchange tube bundle or plate heat exchange structure. The volume and capacity are determined according to the scale of the cold storage.

[0099] The heat exchange unit includes a cold storage side heat exchanger and a chilled water (or refrigerant) flow path, which is used for heat exchange when the medium is charged and discharged.

[0100] Valves and pumps, including charging valve groups and discharging valve groups, respectively control the inflow and outflow paths of chilled water; circulating pumps drive chilled water to circulate between the heat pump, cold storage load and cold storage tank.

[0101] Thermal insulation layer: the tank body is wrapped with high-efficiency thermal insulation material to reduce self-heating loss.

[0102] Optionally, the energy modeling submodule adopts a multivariable dynamic modeling method to establish a cold storage temperature change equation, a cold storage energy change equation and a comprehensive power balance relationship for predicting system state transition.

[0103] Specifically, the core purpose of energy modeling is to build a coupled dynamic relationship model between multiple energy links in the cold storage system (such as photovoltaic power, air-energy heat pumps, power loads, and cold storage devices) to provide prediction input and state expression for subsequent scheduling optimization. It specifically includes the following submodules:

[0104] 1. Photovoltaic power generation model

[0105] Used to predict photovoltaic power output per unit time, the formula is:

[0106] P pv (t) = η pv ·A pv ·G(t)·(1-γ(Tc (t)-25))

[0107] Among them, η pv : Photovoltaic conversion efficiency;

[0108] A pv : PV module area;

[0109] G(t): Solar radiation intensity per unit time (W / m 2 );

[0110] T c (t): component temperature;

[0111] γ: temperature correction coefficient (approximately 0.004 / °C);

[0112] This model dynamically reflects the impact of ambient light and temperature changes on photovoltaic power generation and is a key input for predicting the system's power supply capacity.

[0113] 2. Air source heat pump energy consumption model

[0114] Used to estimate the electrical power required for the air source heat pump to operate according to the outside temperature and cooling load. The formula is:

[0115]

[0116] The COP can be modeled as a function of the external temperature:

[0117] COP(T)=ab·T amb (t)

[0118] Among them, Q cool (t): cooling capacity demand (kW);

[0119] COP: Energy efficiency ratio, which decreases as the outside temperature decreases;

[0120] a, b: coefficients based on the measured data fitting;

[0121] The model captures the “performance degradation” of air source heat pumps and is used to dynamically determine whether they are superior to direct electrical cooling or thermal storage strategies.

[0122] 3. Cold storage load model

[0123] Model the cooling capacity required for the cold storage based on the difference between ambient temperature and set temperature:

[0124] Q load (t)=U·A wall ·(T in -T set )+Q open (t)+Q goods (t)

[0125] Where, U: heat transfer coefficient of the enclosure structure;

[0126] A wall : total wall area;

[0127] T in : Current temperature outside the warehouse;

[0128] T set : Set temperature in the storage;

[0129] Q open (t),Q goods (t): Instantaneous thermal disturbance load such as door opening and cargo loading;

[0130] The model provides basic input for forecasting future loads and helps to adjust the "timeliness" of cooling source supply.

[0131] 4. Cold storage model

[0132] The state transition equation is used to model the energy storage / discharge state of the cold storage tank:

[0133]

[0134] Among them, E cs (t): cold storage surplus energy (kWh);

[0135] P ch ,P dis : Cold storage charging and discharging power;

[0136] η ch ,η dis : Charging and discharging cooling efficiency;

[0137] Δt: time step;

[0138] This model reflects the dynamic behavior of cold storage as an "adjustable cold source" and is a key flexible regulation resource in system optimization.

[0139] 5. Comprehensive power balance constraint model

[0140] Used to ensure closed-loop balance between electrical power and heat load in the system:

[0141] Power balance:

[0142] P pv (t)+P grid (t) = P ashp (t)+P ch (t)-P dis (t)

[0143] Cooling load balancing:

[0144] Qload (t) = Q ashp (t)+Q cs (t)

[0145] Among them, P grid (t) is the power purchased by the power grid; Q cs (t) = P dis (t)·η dis Indicates the amount of cold storage and release; Q ashp (t) represents the cooling capacity released by the air source heat pump.

[0146] This model ensures that the cooling capacity at each moment meets the cold storage load and that energy scheduling is completed within the energy supply boundary, which is a constraint condition of the scheduling algorithm.

[0147] The intelligent scheduling submodule, based on the reinforcement learning algorithm, inputs environmental status information and outputs an optimized energy distribution control strategy to minimize the system energy consumption cost.

[0148] Optionally, the state space is set as:

[0149] S t =[G(t),T amb (t),T in (t),Q load (t),E cs (t),P pv (t),TimeOfDay]

[0150] Where: G(t) is the solar irradiance, T amb (t) is the ambient temperature, T in (t) is the internal temperature of the cold storage, Q load (t) is the cooling load forecast, E cs (t) is the current cold storage energy, P pv (t) is the photovoltaic power output, and TimeOfDay is the current time information.

[0151] Optionally, the action space includes the following three continuous control variables:

[0152] A t =[u ashp ,u ch ,u dis ]

[0153] Among them, the air source heat pump operating power control variable u ashp , the range is [0,1];

[0154] Cold storage charging control variable u ch , the range is [0,1];

[0155] Cooling energy storage control variable u dis , the range is [0,1].

[0156] State transition function S t+1 =f(S t ,A t ), specifically:

[0157]

[0158] The action affects the cooling input, changes the cooling state and the cold storage temperature;

[0159] The cooling load satisfaction is obtained through physical equations;

[0160] Power purchased from the power grid P grid (t) is the energy cost basis.

[0161] Optionally, the reward function of the reinforcement learning algorithm is the negative value of the total electricity cost of the system.

[0162] Specifically, the reward function adopts the energy cost minimization objective:

[0163] R t =-C grid (t)=-P grid (t)·Price(t)·Δt

[0164] Meaning: The reward for each step is a "negative electricity cost." The model will minimize long-term electricity costs by learning to flexibly dispatch heat pumps, cold storage, and photovoltaic utilization under peak and valley electricity prices. Simplifying the reward helps the strategy converge quickly and clearly align with the goal.

[0165] Optionally, the reinforcement learning algorithm of the intelligent scheduling submodule adopts deep deterministic policy gradient (DDPG), and its reinforcement learning process is specifically as follows:

[0166] Initialize the strategy network and value network;

[0167] Initialize the experience replay pool and target network;

[0168] For each task cycle:

[0169] a. The environment perception module gives the current state space S t ;

[0170] b. Policy network output action space A t ;

[0171] c. Calculate the next state S through the energy modeling submodule t+1 , get the reward function R t ;

[0172] d. Storage (S t ,A t ,R t ,S t+1 );

[0173] e. Sample batches from the experience pool for gradient update;

[0174] Use experience replay and target network training strategies;

[0175] Deploy after training until the reward function converges.

[0176] The remote monitoring and feedback module is used to access the cloud platform to achieve remote operation management, data visualization and energy efficiency analysis.

[0177] Optionally, remotely set temperature set points and energy efficiency policy constraints; display current energy consumption structure, cold storage status and scheduling logs; upload data to the cloud platform, support historical data backtracking and energy efficiency evaluation report output.

[0178] Specifically, the remote monitoring and feedback module is the human-machine interaction and operation and maintenance hub of this system. Its core function is to upload on-site operation data to the cloud in real time and provide remote monitoring, parameter distribution, energy efficiency evaluation, and alarm diagnosis services to operation and maintenance personnel and managers through a visual interface and intelligent analysis tools. This module mainly includes the following subsystems and functions:

[0179] 1. System architecture and communication protocol

[0180] The edge gateway is deployed on-site and communicates with various functional modules (environmental sensing, photovoltaic inverter, heat pump PLC, cold storage controller) through industrial buses or Ethernet interfaces such as Modbus-TCP, CAN, RS-485 / RS-232, EtherNet / IP, etc. It performs preliminary filtering, format standardization and encryption on the collected raw data, and then conducts two-way communication with the cloud through MQTT or HTTPS protocols.

[0181] Cloud platform services, based on a containerized microservices architecture, provide device access management, data storage (time series database), visualization screen, alarm rule engine, and energy efficiency analysis model;

[0182] Supports RESTful API and WebSocket to meet the needs of third-party system docking.

[0183] 2. Data Collection and Synchronization

[0184] Real-time data, collecting key operating parameters G(t), T every 1-5 minutes amb (t),T in (t),Q load (t),Ecs (t),P pv (t),TimeOfDay,P grid (t),P ashp (t),P ch (t), etc., and report through the edge gateway;

[0185] Historical archiving: Cloud-based time series databases (such as InfluxDB and TimescaleDB) store all raw and derived data, enabling fast query and export at day / hour / minute granularity.

[0186] Caching and fault tolerance: The edge gateway locally caches at least 24 hours of data when the network is interrupted, and automatically retransmits it after the network is restored to ensure data integrity.

[0187] 3. Visual interface

[0188] See attached Figure 2 The system's visual interface, large-screen monitoring, and system overview display "photovoltaic power generation - heat pump power consumption - grid power purchase curve," as well as the current temperature data of the cold storage.

[0189] 4. Remote parameter distribution and policy adjustment

[0190] Parameter settings include temperature set point, upper and lower limits of cold storage.

[0191] With real-time command issuance, operation and maintenance personnel can adjust key parameters online, such as temperature set points, upper and lower limits of cold storage, etc.; commands are issued through a two-way confirmation mechanism (ACK / NACK) to ensure accurate on-site execution.

[0192] 5. Alarm reminder

[0193] Fault diagnosis automatically identifies faults such as cold storage temperature, cold storage status, heat pump power supply, sensor drift, communication interruption, etc., and issues alarm reminders.

[0194] Beneficial effects

[0195] Operation and maintenance efficiency is greatly improved: through visualization and automatic alarms, operation and maintenance personnel can quickly locate faults and optimization space; flexible response strategy adjustment: algorithms and operating parameters can be remotely distributed in real time, achieving "second-level" strategy iteration; data-driven decision-making: historical data and predictive analysis are combined to provide a scientific basis for future expansion and upgrades; safe and reliable: industrial-grade communication and encryption protocols are used, and edge cache disaster recovery design is designed to ensure continuous and stable operation of the system.

[0196] A cold storage device adopts any of the above cold storage systems.

[0197] Compared with the existing technology, the above technical solution has the following beneficial effects:

[0198] 1. Significantly reduce energy costs. Reinforcement learning algorithms are used to schedule heat pumps, cold storage, and photovoltaic output in real time, prioritizing the use of low-price electricity and self-generated photovoltaic power, reducing peak-period electricity purchases and minimizing overall system electricity costs.

[0199] 2. Improve the self-consumption rate of solar energy. The environmental sensing module monitors irradiance and power generation in real time. Combined with intelligent scheduling, it automatically stores energy or directly drives loads during periods of sufficient sunlight, maximizing the utilization of photovoltaic power generation and avoiding wasteful solar power.

[0200] 3. Peak shaving and valley filling, load smoothing. The cold storage module stores cold energy during off-peak hours or when there is surplus photovoltaic power, and releases it during peak hours, effectively smoothing fluctuations in heat pump and grid loads, improving grid access stability and equipment lifespan.

[0201] 4. Efficient refrigeration and stable temperature control. Using a cold storage thermal model, the system meets cooling load requirements in real time and maintains temperature deviation to ensure product quality.

[0202] 5. Dynamic self-adaptation and multi-objective optimization. Based on deep reinforcement learning, the system can self-learn peak and valley electricity prices, weather, and load variations, achieving multi-objective optimization control of energy costs, carbon emissions, and equipment lifespan. The strategy is continuously optimized as the environment evolves.

[0203] 6. Modular design, easy to expand and integrate. Each functional module (environmental sensing, photovoltaic power generation, heat pump, cold storage, energy efficiency modeling, remote monitoring) is highly decoupled, supporting on-demand addition, subtraction, or upgrade, facilitating integration into cold storage, warehousing, and industrial refrigeration scenarios of varying sizes.

[0204] 7. Convenient remote monitoring and operation and maintenance. Cloud platform access enables real-time data visualization, remote parameter distribution, and fault warnings, greatly improving system operation and maintenance efficiency and management transparency. It supports multi-terminal access from mobile devices and industrial touch terminals.

[0205] 8. Reduce carbon emissions and achieve green and low-carbon development. Replacing traditional electric heating / cooling with solar energy and high-efficiency heat pumps achieves synergy between clean energy and energy storage, significantly reducing the carbon footprint of system operation and meeting energy conservation, emission reduction, and carbon neutrality requirements.

[0206] Through the above beneficial effects, the present invention can provide efficient, reliable and sustainable intelligent energy efficiency management solutions for the fields of cold chain, preservation, low-temperature processing, etc. at both technical and economic levels.

[0207] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. An energy-saving refrigeration cold storage system, characterized in that: include: Environmental sensing module, used to collect real-time environmental and operating parameters such as solar irradiance, ambient temperature, internal temperature of cold storage, cooling load demand, and photovoltaic power generation output; Solar power generation module, used to convert solar radiation into DC or AC electricity for system use; Air energy heat pump module, used to provide cooling or heating energy according to control instructions; The cold storage module is used to store cold energy during off-peak hours or when there is sufficient solar energy, and release the cold energy during high-load hours; Energy efficiency modeling and intelligent control module, which includes: Energy modeling submodule, used to establish a multi-source energy flow mathematical model including photovoltaic power generation, cold storage thermal characteristics, air energy heat pump efficiency, and the charging and discharging dynamics of cold storage units; The intelligent scheduling submodule, based on the reinforcement learning algorithm, inputs environmental status information and outputs an optimized energy distribution control strategy to minimize the system energy consumption cost; The remote monitoring and feedback module is used to access the cloud platform to achieve remote operation management, data visualization and energy efficiency analysis.

2. The cold storage system according to claim 1, characterized in that: The air energy heat pump module mainly includes a compressor, an evaporator, a condenser, an expansion valve, a fan and a water pump, and a controller.

3. The cold storage system according to claim 1, characterized in that: The cold storage module includes a cold storage tank, a phase-change cold storage medium, and a heat exchange unit.

4. The cold storage system according to claim 1, characterized in that: The energy modeling submodule adopts a multivariable dynamic modeling method to establish a cold storage temperature change equation, a cold storage energy change equation, and a comprehensive power balance relationship to predict system state transitions.

5. The cold storage system according to claim 4, characterized in that: The state space set by the reinforcement learning algorithm is: S t =[G(t),T amb (t),T in (t),Q load (t),E cs (t),P pv (t),TimeOfDay] Where: G(t) is the solar irradiance, T amb (t) is the ambient temperature, T in (t) is the internal temperature of the cold storage, Q load (t) is the cooling load forecast, E cs (t) is the current cold storage energy, P pv (t) is the photovoltaic power output, and TimeOfDay is the current time information.

6. The cold storage system according to claim 5, characterized in that: The action space set by the reinforcement learning algorithm includes the following three continuous control variables: Air source heat pump operating power control variables; Cold storage charging control variables; Cold storage and release can control variables.

7. The cold storage system according to claim 6, characterized in that: The reward function of the reinforcement learning algorithm is the negative value of the total electricity cost of the system.

8. The cold storage system according to claim 7, characterized in that: The reinforcement learning algorithm of the intelligent scheduling submodule adopts deep deterministic policy gradient (DDPG), and its reinforcement learning process is as follows: Initialize the strategy network and value network; Initialize the experience replay pool and target network; For each task cycle: a. The environment perception module gives the current state space S t ; b. Policy network output action space A t ; c. Calculate the next state S through the energy modeling submodule t+1 , get the reward function R t ; d. Storage (S t ,A t ,R t ,S t+1 ); e. Sample batches from the experience pool for gradient update; Use experience replay and target network training strategies; Deploy after training until the reward function converges.

9. The cold storage system according to claim 1, characterized in that: Remotely set temperature set points and energy efficiency strategy constraints; display current energy consumption structure, cold storage status and scheduling logs; upload data to the cloud platform, support historical data backtracking and energy efficiency evaluation report output.

10. A cold storage device, comprising the cold storage system according to any one of claims 1 to 9.

Images