A household energy double closed loop optimization scheduling method and intelligent power control device

Abstract

The present application relates to the technical field of home energy management. The present application discloses a kind of home energy double closed loop optimization scheduling method and intelligent power control device, it includes: S1 multi-source data acquisition preprocessing, obtains the operating data of equipment such as power grid, photovoltaic through metering module, synchronously obtains dynamic electricity price and VPP scheduling instruction;S2 based on GA-LP hybrid model, with the lowest cost, VPP highest income as target generates equipment operation plan;S3 adopts MPC technology revision plan, executes anti-backflow, isolated network standby power and VPP instruction response;S4 gathers feedback data, strengthens learning and updates model and optimizes load priority weight.Double closed loop architecture combines GA-LP and MPC algorithm, and the scheduling accuracy is improved by more than 30%.Multi-modal communication plus external WiFi antenna, guarantee complex environment transmission stability;Hard metering chip plus upstream collection, realize accurate measurement;Anti-backflow is less than or equal to 2s, standby power switching is less than or equal to 200ms, and response speed leads.

Description

Technical Field

[0001] This invention relates to the field of home energy management technology, and in particular to a home energy dual-closed-loop optimized scheduling method and intelligent power control device. Background Technology

[0002] With the popularization of distributed energy and electrical equipment such as photovoltaics, energy storage, electric vehicle charging piles, and smart home appliances, the home energy system has been upgraded from a single power terminal to an integrated ecosystem of "production-storage-charging-use".

[0003] In existing technologies, some solutions attempt to optimize dynamic electricity price response through a two-layer optimization model, but fail to take into account the actual operating constraints of household appliances; some VPP scheduling methods, while considering load priority, lack closed-loop linkage with real-time metering data.

[0004] Existing technologies suffer from several problems: First, the scheduling dimension is too narrow. Most solutions focus only on peak-valley arbitrage or single-device control, failing to achieve deep coordination between dynamic electricity pricing, VPP scheduling commands, and the status of multiple household devices, resulting in insufficient energy cost optimization. Second, communication compatibility and stability are inadequate. Single communication methods cannot meet the interface requirements of multiple devices such as inverters, charging piles, and heat pumps, and the metal shielding of distribution boxes easily leads to wireless communication interruptions, affecting control command transmission. Third, metering integrity and accuracy are lacking. The use of soft metering or downstream acquisition methods fails to account for the controller's own power consumption, leading to deviations in electricity purchase and supply data, failing to meet electricity metering standards. Fourth, control response is lagging. Response times are too long in critical scenarios such as reverse current prevention and grid outage backup switching, failing to meet regional policy requirements and ensuring power continuity. Fifth, VPP response capabilities are weak. Only energy storage batteries are included in the scheduling scope, without considering the adjustment potential of controllable loads, resulting in limited dispatchable capacity and difficulty in maximizing the response to VPP commands. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention discloses an optimized scheduling scheme with multi-device collaboration capabilities, rapid response characteristics, high-precision metering, and wide-area communication compatibility, which solves the comprehensive pain points of home energy management. It is a dual closed-loop optimized scheduling method and intelligent power control device for home energy.

[0006] This invention discloses a dual-closed-loop optimization scheduling method for household energy, which includes the following steps:

[0007] S1: Multi-source data acquisition and preprocessing: Real-time operation data of power grid, photovoltaic, energy storage and household electrical equipment are acquired through the metering module, while dynamic electricity price data and virtual power plant (VPP) dispatch instructions are obtained.

[0008] S2: Outer Global Optimization Scheduling: Based on a hybrid optimization model of genetic algorithm and linear programming, with the goal of minimizing energy costs and maximizing VPP response benefits, it generates future equipment operation plans by taking into account dynamic electricity price curves, VPP scheduling instructions, equipment parameters and user needs.

[0009] S3: Inner Real-Time Control Closed Loop: Employs model predictive control technology to dynamically correct the outer plan based on real-time data, and executes anti-backflow control, islanded backup power control, and VPP command response;

[0010] S4: Status Feedback and Adaptive Adjustment: Collect device status and command execution results, update and optimize model parameters through reinforcement learning algorithms, and optimize load priority weights based on user electricity consumption behavior data.

[0011] Furthermore, in step S1, the real-time operating data includes the controller's own power consumption;

[0012] The collected data undergoes digital filtering, harmonic correction, and outlier removal.

[0013] Dynamic electricity price data includes day-ahead spot electricity prices and intraday real-time price correction data at 15-minute intervals.

[0014] Furthermore, in step S2, the equipment operation plan includes the charging and discharging periods and power of the energy storage battery, the start-up time of the electric vehicle charging pile, and the start-up and shutdown periods of controllable loads including heat pumps and smart home appliances.

[0015] Furthermore, in step S3, the anti-reverse flow control controls the inverter output power through a proportional-integral-derivative PID regulation algorithm;

[0016] The islanded backup power control disconnects low-priority load circuits in sequence according to the energy storage state of charge (SOC) and the preset load priority sequence to ensure power supply to critical loads.

[0017] VPP command response decomposes scheduling commands into energy storage charging and discharging power and controllable load adjustment, and dynamically allocates adjustment resources.

[0018] Furthermore, data transmission is achieved through a multi-mode communication network that supports Ethernet, WiFi, 4G, RS485, LAN, and BPLC power line carrier. Among these, WiFi uses an external antenna to ensure that the communication distance is not less than 2 meters when the distribution box is closed.

[0019] This invention discloses an intelligent power control device for any of the above-mentioned dual closed-loop optimization scheduling methods for household energy, comprising:

[0020] The core control module is used to run the dual closed-loop optimization scheduling algorithm;

[0021] The metering module is used to collect end-to-end power data, including the device's own power consumption.

[0022] A multimodal communication module is used for data interaction with the power grid platform, VPP platform, and household electrical appliances;

[0023] Interface and structure modules are used to provide device access interfaces and meet physical installation and environmental protection requirements.

[0024] Furthermore, the metering module is equipped with a hard metering chip, supporting 1P2W, 3P3W or 3P4W wiring methods, with active power metering accuracy of no less than level 1 and reactive power metering accuracy of no less than level 2.

[0025] The metering current transformer (CT) is located upstream of the device's power supply node and includes both built-in and external CTs; the module is also equipped with a hardware clock with temperature compensation function.

[0026] Furthermore, the multimodal communication module includes:

[0027] The northbound communication unit supports Ethernet, external antenna WiFi, and optional 4G communication, and is compatible with MQTT, IEC-61870-5-104, and OpenADR protocols.

[0028] The southbound communication unit includes multiple RS485 interfaces, a LAN port, and an optional BPLC carrier interface, supporting Modbus-RTU / TCP protocols;

[0029] The near-end communication unit adopts WiFi access point (AP) mode for local parameter configuration.

[0030] Furthermore, the interface and structural modules adopt DIN35 rail mounting, the housing width is no more than 108 mm, and the protection level is IP30;

[0031] Provides digital input / output (DI / DO), 12-volt DC input / output, and pulse output interfaces, supporting wires with a cross-sectional area of ​​not less than 35 square millimeters;

[0032] The outer shell is made of flame-retardant polycarbonate (PC) with added glass fiber, and the operating temperature range is -25 degrees Celsius to 70 degrees Celsius.

[0033] Furthermore, it also includes software functional modules, which at least include:

[0034] The module includes a data acquisition and preprocessing module, a dual-closed-loop scheduling module integrating an outer layer genetic algorithm-linear programming (GA-LP) optimization and an inner layer model predictive control (MPC) algorithm, an equipment management module for managing inverters, charging piles and controllable loads, a VPP coordination module for parsing and responding to VPP commands, a dynamic electricity price management module for managing dynamic electricity price data, and an alarm and log module for recording operating status and anomalies.

[0035] The beneficial effects of this invention are:

[0036] 1. More comprehensive scheduling optimization: The dual closed-loop architecture combines GA-LP and MPC algorithms to achieve an organic unity between long-cycle planning and short-cycle correction, improving the accuracy of dynamic electricity price response and VPP collaborative scheduling by more than 30%.

[0037] 2. Stable and reliable communication: The multi-mode communication network covers the interface requirements of various devices, and the external WiFi antenna design effectively overcomes the signal attenuation problem caused by the metal shielding of the distribution box, significantly improving the communication coverage and data transmission reliability in complex home environments.

[0038] 3. Accurate and complete metering: With a hard metering chip and upstream acquisition design, taking into account the controller's own power consumption, the active power error is ≤±0.5%, meeting the IEC62053-21 / 23 standard, and significantly improving data reliability;

[0039] 4. Leading response speed: Anti-reverse current response ≤2s, islanded grid backup power switching ≤200ms, far exceeding the current technology level, ensuring power safety and continuity;

[0040] 5. Enhanced VPP Response Capability: Incorporating controllable loads into the scheduling scope increases scheduling capacity by over 50%, maximizing the response to VPP commands and generating additional benefits;

[0041] 6. High compatibility and reliability: The rail-mounted design is compatible with household electrical distribution boxes. The wide temperature range design and flame-retardant materials ensure a service life of more than 8 years and meet the requirements for long-term stable operation. Attached Figure Description

[0042] Figure 1 This is a flowchart of the dual closed-loop optimization scheduling method for household energy in the embodiments of this application.

[0043] Figure 2 This is a schematic diagram of the hardware architecture of the intelligent power controller in the embodiments of this application.

[0044] Figure 3 This is a schematic diagram of the core logic of the dual closed-loop scheduling software in the embodiments of this application. Detailed Implementation

[0045] To enable those skilled in the art to better understand the present invention, the technical solutions in the specific embodiments of the present invention will be clearly and completely described below.

[0046] This invention discloses a dual-closed-loop optimization scheduling method for household energy, which includes the following steps:

[0047] S1: Multi-source data acquisition and preprocessing: Real-time operation data of power grid, photovoltaic, energy storage and household electrical equipment are acquired through the metering module, while dynamic electricity price data and virtual power plant (VPP) dispatch instructions are obtained.

[0048] S2: Outer Global Optimization Scheduling: Based on a hybrid optimization model of genetic algorithm and linear programming, with the goal of minimizing energy costs and maximizing VPP response benefits, it generates future equipment operation plans by taking into account dynamic electricity price curves, VPP scheduling instructions, equipment parameters and user needs.

[0049] S3: Inner Real-Time Control Closed Loop: Employs model predictive control technology, dynamically corrects the outer plan based on real-time data, executes anti-backflow control, islanded backup power control, and VPP command response. The anti-backflow control response time is no more than 2 seconds, and the islanded backup power switching response time is no more than 200 milliseconds.

[0050] S4: Status Feedback and Adaptive Adjustment: Collect device status and command execution results, update and optimize model parameters through reinforcement learning algorithms, and optimize load priority weights based on user electricity consumption behavior data.

[0051] In this invention, S1 multi-source data acquisition and preprocessing achieves full-dimensional information coverage, simultaneously acquiring equipment operation data, dynamic electricity prices, and VPP dispatch instructions. Combined with data optimization processing, it provides data support for subsequent dispatching, ensuring the integrity and reliability of decision-making. S2 employs a hybrid optimization model combining genetic algorithms and linear programming, with a dual-objective orientation that balances energy cost control and VPP response benefits. By integrating multiple input parameters, it generates a scientific equipment operation plan, achieving global-level dispatch configuration. S3 relies on model predictive control technology to complete real-time dynamic correction, specifically executing anti-backflow, islanded backup power, and VPP command response control. Clear response time constraints ensure rapid response in critical scenarios, effectively avoiding backflow risks and ensuring power continuity during grid outages. S4 achieves adaptive adjustment through reinforcement learning algorithms and user electricity behavior analysis, dynamically updating model parameters and load priority weights, allowing the dispatch scheme to continuously adapt to changes in equipment status and user habits, improving long-term operational economy and adaptability. The entire set of steps forms a complete closed loop, from data acquisition to plan generation, real-time control, and feedback optimization, progressing layer by layer to achieve refined, intelligent, and efficient home energy dispatching.

[0052] In one implementation, in step S1, the real-time operating data includes the controller's own power consumption; the collected data undergoes digital filtering, harmonic correction, and outlier removal processing; the dynamic electricity price data includes the day-ahead spot electricity price and intraday real-time electricity price correction data at 15-minute intervals.

[0053] Step S1 incorporates the controller's own power consumption into the real-time operational data acquisition range, filling the gap in existing metering technology and ensuring complete power data across the entire chain. This effectively avoids deviations in power purchase and supply data caused by the failure to account for the device's own energy consumption, meeting the requirements of power metering standards. The acquired data undergoes digital filtering, harmonic correction, and outlier removal, effectively filtering noise data caused by grid fluctuations and equipment interference, correcting measurement deviations caused by harmonic distortion, and eliminating interference from sudden abnormal data on dispatching decisions. This improves data accuracy and stability, providing high-quality, highly reliable data support for subsequent outer-layer optimized dispatching and inner-layer real-time control, ensuring the scientific nature of algorithmic decisions. Dynamic electricity price data simultaneously covers the day-ahead spot price and intraday real-time price correction data at 15-minute intervals. This provides a long-term planning basis for the outer-layer 24-hour global optimized dispatching and, through high-frequency intraday price updates, allows the dispatching scheme to respond promptly to short-term price fluctuations, avoiding insufficient energy cost optimization due to delayed price information and improving the economic efficiency of household energy use.

[0054] In one implementation, step S2 includes the charging and discharging periods and power of the energy storage battery, the start-up time of the electric vehicle charging pile, and the start-up and stop periods of controllable loads including the heat pump and smart home appliances.

[0055] In terms of coverage, the operation plan encompasses core equipment such as energy storage batteries, electric vehicle charging stations, heat pumps, and smart home appliances, achieving unified scheduling of all devices across the entire home "production-storage-charging-use" chain. This avoids energy waste or coordination conflicts caused by single-device control, allowing various devices to work together synergistically. Differentiated plans are developed for the operating characteristics of different devices: clearly defining the charging and discharging periods and power of energy storage batteries to ensure efficient matching with dynamic electricity price peak-valley changes; limiting the start-up time of electric vehicle charging stations to avoid peak electricity consumption and reduce costs; and planning the start-up and shutdown periods of controllable loads to achieve a balance between peak-shaving electricity consumption and efficient energy utilization, ensuring that the operation of each device aligns with optimization goals. The plan design directly serves the dual objectives of minimizing energy costs and maximizing VPP response benefits, transforming abstract optimization models into concrete, executable equipment operation instructions. Through multi-device collaborative scheduling, it not only taps into peak-valley arbitrage opportunities but also incorporates charging stations and controllable loads into the VPP response resource pool, expanding dispatchable capacity and ensuring more comprehensive VPP command responses, thereby improving the economy and profitability of home energy management.

[0056] In one implementation method, in step S3, the anti-reverse current control controls the inverter output power through a proportional-integral-derivative PID regulation algorithm; the islanded backup power control disconnects low-priority load circuits in sequence according to the energy storage state of charge (SOC) and a preset load priority sequence to ensure power supply to critical loads; the VPP command response decomposes the scheduling command into energy storage charging and discharging power and controllable load adjustment amount, and dynamically allocates adjustment resources.

[0057] This implementation method employs a mature and reliable proportional-integral-derivative (PI-DE) control algorithm for anti-reverse current control. It can quickly capture changes in grid power and regulate inverter output power, achieving immediate suppression of reverse current trends. This ensures grid operation safety while avoiding energy waste due to insufficient control precision, and its response speed fully meets the requirements of efficient dispatching. Islanded backup power control combines energy storage state of charge (SOC) with load priority sequences. By orderly disconnecting low-priority load circuits, it ensures power supply to critical loads even with limited energy storage capacity. This avoids user inconvenience caused by indiscriminate power outages and ensures power continuity in backup scenarios, solving the problems of chaotic backup power switching and unreliable power supply to critical equipment in traditional systems. VPP command response rationally decomposes dispatch commands, incorporating both energy storage and controllable loads into the regulation resource pool and dynamically allocating regulation tasks. This breaks the limitation of relying solely on energy storage to respond to VPP commands, significantly expanding dispatchable capacity and making VPP command response more efficient and profitable.

[0058] As one implementation method, data transmission is performed through a multi-mode communication network that supports Ethernet, WiFi, 4G, RS485, LAN, and BPLC power line carrier. In this case, WiFi uses an external antenna to ensure that the communication distance is not less than 2 meters when the distribution box is closed.

[0059] This implementation method covers diverse communication methods including Ethernet, WiFi, 4G, RS485, LAN, and power line carrier, fully adapting to the interface requirements of different devices such as power grid platforms, VPP platforms, inverters, charging piles, and smart home appliances. It avoids the limitations of a single communication mode being incompatible with multiple devices, enabling seamless interaction of various data. Addressing the industry challenge of metal casings in distribution boxes easily shielding wireless signals, the WiFi uses an external antenna design, ensuring that the communication distance in a closed state meets actual installation requirements, completely solving the problem of control command transmission failure caused by wireless communication interruptions. The synergistic complementarity of multiple communication methods forms redundancy protection, allowing flexible selection based on the home installation environment and device type, significantly reducing the risk of data transmission packet loss and ensuring real-time synchronization of key information such as dynamic electricity price data, VPP scheduling commands, and real-time device status. A stable and efficient communication link provides a solid foundation for modeling of outer-layer global optimization scheduling and rapid response of inner-layer real-time control, improving the adaptability and operational reliability of the entire energy management system, perfectly meeting the diverse communication needs of complex home energy scenarios.

[0060] This invention discloses an intelligent power control device, comprising: a core control module for running a dual closed-loop optimization scheduling algorithm; a metering module for collecting full-link power data, including the device's own power consumption; a multimodal communication module for data interaction with the power grid platform, VPP platform, and household electrical appliances; and an interface and structure module for providing device access interfaces and meeting physical installation and environmental protection requirements.

[0061] The core control module serves as the core, engaging in bidirectional data exchange with the metering module that collects end-to-end power data, the multimodal communication module that enables data interaction with the power grid platform, VPP platform, and household electrical equipment, and the interface and structural module that provides device access interfaces while meeting physical installation and environmental protection requirements. Simultaneously, it runs software functional modules that integrate data acquisition and preprocessing, dual closed-loop scheduling, and other functions, thus constructing a hardware and software collaborative home energy optimization scheduling and control system.

[0062] This intelligent power control device features a scientifically divided modular design and a functionally coordinated closed-loop system. From computation and data acquisition to communication and installation protection, it constructs a comprehensive support system, adapting to the core needs of dual-loop home energy scheduling and effectively addressing many shortcomings of existing technologies. The core control module, as the device's computational core, provides a stable and efficient operating platform for the dual-loop optimized scheduling algorithm, ensuring smooth integration between the outer-layer global optimization and the inner-layer real-time control. This is the core support for achieving refined home energy scheduling, enabling the realization of dynamic electricity price response and VPP collaborative scheduling goals. The metering module focuses on full-link energy data acquisition, including the device's own power consumption within the acquisition scope. This fills the gap in incomplete metering in traditional solutions, avoiding the impact of data deviations on scheduling decisions, providing a reliable data foundation for algorithm operation, and meeting stringent electricity metering standards. The multi-modal communication module establishes a comprehensive data interaction channel, achieving seamless integration with the power grid platform, VPP platform, and various devices within the home. This solves the problem of insufficient adaptability of a single communication method, ensuring the real-time transmission of key information such as scheduling commands, electricity price data, and device status. The interface and structural modules take into account both device access and installation protection. The rich interfaces meet the access needs of various types of devices, are suitable for the installation scenario of home distribution boxes, and the enhanced environmental protection capabilities ensure stable operation of the device under complex conditions.

[0063] As one implementation method, the metering module is equipped with a hard metering chip, which supports 1P2W, 3P3W or 3P4W wiring methods. Its active power metering accuracy is not lower than Class 1 and reactive power metering accuracy is not lower than Class 2. The metering current transformer (CT) is located upstream of the power supply node of the device, including built-in CT and external CT. The module is also equipped with a hardware clock with temperature compensation function.

[0064] This implementation method incorporates a hard metering chip and clearly defines standards for active power metering accuracy of no less than Class 1 and reactive power metering accuracy of no less than Class 2. Compared to traditional soft metering solutions, it significantly improves the accuracy of power data measurement, effectively avoids metering errors caused by harmonic interference and load fluctuations, fully meets the relevant power metering specifications, and ensures the fairness and accuracy of power purchase and distribution data accounting. It supports multiple wiring methods such as 1P2W, 3P3W, or 3P4W, flexibly adapting to different household power supply types (single-phase and three-phase), requiring no additional adaptation or modification for installation and use, significantly improving the module's versatility and ease of installation. The metering current transformer is located upstream of the device's power supply node, and with the complementary design of built-in and external CTs, it successfully incorporates the controller's own power consumption into the metering range, filling the metering gap of traditional downstream acquisition methods, achieving complete collection of power data across the entire chain, and avoiding scheduling decision deviations due to missing energy consumption data. Equipped with a hardware clock featuring temperature compensation, it can effectively resist the impact of ambient temperature changes on timing accuracy, ensure the stability of time synchronization, provide a time reference for dynamic electricity price data matching, scheduling command execution timing, equipment operation log recording, etc., ensure the timing consistency between scheduling plans and real-time control, and enhance the reliability of the entire energy management system.

[0065] As one implementation, the multimodal communication module includes: a northbound communication unit that supports Ethernet, external antenna WiFi, and optional 4G communication, and is compatible with MQTT, IEC-61870-5-104, and OpenADR protocols; a southbound communication unit that includes multiple RS485 interfaces, a LAN port, and an optional BPLC carrier interface, and supports Modbus-RTU / TCP protocols; and a near-end communication unit that adopts WiFi access point (AP) mode for local parameter configuration.

[0066] This implementation's northbound communication unit supports Ethernet, external antenna WiFi, and optional 4G communication, compatible with multiple mainstream protocols. It can seamlessly connect to the power grid platform and VPP platform, ensuring efficient transmission of critical information such as dynamic electricity price data and dispatch instructions. The external antenna design completely eliminates the influence of the metal shielding of the distribution box, ensuring the stability of wireless communication. The southbound communication unit is equipped with multiple interfaces and optional carrier interfaces, compatible with common protocols, and can fully adapt to the access needs of different types of devices such as inverters, charging piles, and smart home appliances within the home. It avoids the limitation of a single interface not being compatible with multiple devices, enabling centralized management and data interaction of home energy devices. The near-end communication unit adopts a WiFi access point mode, providing a convenient channel for local parameter configuration. Device debugging and parameter modification can be completed without complex wiring, significantly improving installation and maintenance efficiency. The three units work together to form a complete communication link from upper-level platform connection and home device linkage to local operation and maintenance configuration. It balances the stability of remote data transmission with the convenience of local operation, ensuring that all kinds of communication needs are met, and providing solid communication support for the smooth operation of dual closed-loop optimized dispatch.

[0067] As one implementation method, the interface and structural module adopts DIN35 rail mounting, the housing width is no more than 108 mm, and the protection level is IP30; it provides digital input / output (DI / DO), 12-volt DC input / output and pulse output interfaces, and supports wires with a cross-sectional area of ​​not less than 35 square millimeters; the housing material is polycarbonate PC with added glass fiber flame retardant material, and the operating temperature range is -25 degrees Celsius to 70 degrees Celsius.

[0068] This implementation uses a DIN35 rail mounting method and has a reasonably narrow housing width, allowing for rapid deployment without additional modifications to the distribution box. This significantly reduces installation difficulty and space requirements, making it suitable for compact installation environments in home settings. With an IP30 protection rating, it effectively prevents dust intrusion, protecting internal components from contamination and ensuring long-term stable operation.

[0069] The rich interface configuration fully meets the needs of multi-device access. Digital input / output, 12V DC input / output, and pulse output interfaces cover the control and signal transmission needs of various devices. The design supports large cross-sectional area wire access, ensuring a firm and reliable connection with high-power devices and avoiding poor contact or safety hazards caused by wire compatibility issues. The shell is made of flame-retardant polycarbonate with added glass fiber, which combines high strength and fire resistance, effectively resisting physical impacts and fire risks in daily use. The wide temperature range design breaks the limitations of ambient temperature, maintaining stable operation in both extremely cold and extremely hot conditions, adapting to different regional climates. The overall design balances practicality and reliability, allowing the device to meet diverse device access needs while possessing excellent installation adaptability, safety protection capabilities, and environmental tolerance, providing solid structural support for the long-term stable operation of the dual closed-loop energy dispatching system in the home.

[0070] As one implementation method, it also includes software functional modules, which at least include: a module for data acquisition and preprocessing; a dual closed-loop scheduling module integrating an outer layer genetic algorithm-linear programming (GA-LP) optimization and an inner layer model predictive control (MPC) algorithm; an equipment management module for managing inverters, charging piles, and controllable loads; a VPP coordination module for parsing and responding to VPP commands; a dynamic electricity price management module for managing dynamic electricity price data; and an alarm and log module for recording operating status and anomalies.

[0071] This implementation's data acquisition and preprocessing module filters interference and corrects deviations, ensuring the accuracy of data across the entire chain and providing a high-quality data foundation for subsequent optimization algorithms and control logic. The dual-closed-loop scheduling module integrates an outer-layer genetic algorithm-linear programming optimization with an inner-layer model predictive control algorithm, achieving an organic unity of long-cycle global planning and short-cycle real-time correction, improving energy economy and response benefits. The equipment management module enables centralized management and control of inverters, charging piles, and controllable loads, breaking down barriers to multi-device collaboration and providing efficient equipment-level support for multimodal scheduling. The VPP collaboration module is specifically responsible for instruction parsing, resource allocation, and result feedback, strengthening its linkage with the VPP platform and ensuring more effective execution of scheduling instructions. The dynamic electricity price management module updates and stores electricity price data in real time, providing timely and effective price basis for the dual-layer optimization framework, ensuring peak-valley arbitrage and cost optimization effects. The alarm and log module comprehensively records operating status and abnormal situations, providing clear guidance for equipment maintenance and problem troubleshooting, significantly improving the reliability and maintainability of system operation. Each module performs its own function while working closely together to build a complete software system covering data processing, scheduling decisions, equipment management, platform linkage, and operation and maintenance support, ensuring the efficient implementation of the dual closed-loop scheduling method.

[0072] The technical solution is as follows:

[0073] 1. A Dual-Closed-Loop Optimization Scheduling Method for Household Energy

[0074] This method achieves refined scheduling of home energy through a dual closed-loop architecture of outer-layer global optimization and inner-layer real-time control. Figure 1 The flowchart for the dual closed-loop optimization scheduling method for household energy is as follows:

[0075] S1: Multi-source data acquisition and preprocessing

[0076] 1) Real-time data such as grid voltage / current, photovoltaic power generation, energy storage SOC (State of Charge), and power consumption of each device are collected through a high-precision metering module, including the power consumption of the controller itself;

[0077] 2) After digital filtering, harmonic correction, and outlier removal, the data is synchronized to the local computing unit and the cloud platform;

[0078] 3) Obtain Day-Ahead spot electricity price and 15-minute Intraday real-time electricity price correction data, as well as adjustment instructions and response requirements from the VPP dispatch platform, through northbound communication.

[0079] S2: Outer layer global optimization scheduling (24-hour planning layer)

[0080] 1) Based on the genetic algorithm-linear programming (GA-LP) hybrid optimization technique, an optimization model is constructed with the goal of minimizing energy costs and maximizing VPP response benefits;

[0081] 2) Input parameters include dynamic electricity price curve, VPP dispatch instructions, equipment parameters (energy storage capacity, charging pile power, photovoltaic predicted power generation), and user settings (backup power priority, charging demand).

[0082] 3) Output a 24-hour equipment operation plan, including the energy storage charging and discharging period and power, the charging pile start-up time, and the start-up and shutdown periods of controllable loads (heat pumps, smart home appliances).

[0083] S3: Inner layer real-time control closed loop (15-minute correction layer)

[0084] 1) Model predictive control (MPC) technology is adopted to adjust the outer plan every 15 minutes based on real-time metering data, equipment status feedback and the latest electricity price correction value;

[0085] 2) Anti-backflow control: Real-time monitoring of grid power; when a backflow trend is detected, the inverter output power is controlled by a proportional-integral-derivative (PID) adjustment algorithm, with a response time of ≤2 seconds.

[0086] 3) Isolated grid backup power control: The grid power failure detection response is ≤200ms. Based on the energy storage SOC and load priority sequence, low priority circuits are disconnected in sequence to ensure power supply to critical loads.

[0087] 4) VPP command response: Decompose the VPP scheduling command into energy storage charging and discharging power and controllable load adjustment amount, and dynamically allocate adjustment resources in combination with load priority to expand the scheduling capacity.

[0088] S4: Status Feedback and Adaptive Adjustment

[0089] 1) Collect equipment operating status and scheduling command execution results in real time, and update and optimize model parameters through reinforcement learning algorithms;

[0090] 2) Optimize load priority weights based on user electricity consumption behavior data to improve scheduling economy while meeting comfort requirements.

[0091] 2. Intelligent power control device (hardware and software modules for implementing the above method)

[0092] The device includes a hardware architecture and software functional modules. The hardware provides the basic support for the method implementation, while the software modules execute the scheduling algorithm logic.

[0093] Hardware architecture

[0094] Figure 2 The following is a schematic diagram of the hardware architecture of the intelligent power controller, with the specific modules as follows:

[0095] 1) Core control module: It adopts a high-performance core board, with more than 1GB of RAM and more than 8GB of ROM, providing high-efficiency computing power and supporting the operation of dual closed-loop scheduling algorithm;

[0096] 2) High-precision metering module: Equipped with a hard metering chip, supporting 1P2W / 3P3W / 3P4W wiring methods, active power accuracy ≥ Class 1, reactive power accuracy ≥ Class 2; CT acquisition position is located upstream of the product power supply node, including built-in CT (100A) and external CT (250A), realizing its own power consumption calculation; Equipped with temperature-compensated hardware clock, the clock is maintained for ≥ 5 years after power failure;

[0097] 3) Multimodal communication module: Northbound communication supports Ethernet (10 / 100Mbps), WiFi (2.4GHz, external antenna shielded), and 4G (optional), and is compatible with MQTT, IEC-61870-5-104, and OpenADR protocols, supporting automatic switching and manual configuration processes; Southbound communication includes 3 RS485 ports, 1 LAN port, and an optional BPLC carrier interface, supporting Modbus-RTU / TCP protocols and adapting to multiple device access; Near-end communication adopts WiFiAP mode, with a communication distance of ≥2 meters when the power distribution box is closed, meeting local parameter configuration requirements;

[0098] 4) Interface and structural module: DIN35 rail mounting, width ≤108mm, IP30 protection rating; equipped with DI / DO, 12V input / output, pulse output, etc., supporting ≥35mm... 2 Wire connection; the outer shell is made of PC + glass fiber flame retardant material, which meets the working environment requirements of -25℃-70℃.

[0099] Software functional modules

[0100] Figure 3 This is a schematic diagram of the core logic of the dual closed-loop scheduling software. The specific implementation logic is as follows:

[0101] 1) Data acquisition and preprocessing module: This module acquires, filters, and corrects data such as voltage, current, and power to ensure data accuracy.

[0102] 2) Dual closed-loop scheduling module: integrates the outer GA-LP optimization algorithm and the inner MPC real-time control algorithm to execute the core scheduling logic;

[0103] 3) Equipment Management Module: Supports the access and status monitoring of ≥5 inverters, ≥2 charging piles, heat pumps, smart loads, and other equipment;

[0104] 4) VPP Collaboration Module: Implements VPP command parsing, resource allocation adjustment, and response result feedback; supports identity authentication and data encryption.

[0105] 5) Dynamic electricity price management module: Stores and updates dynamic electricity price data, supporting the operation of the two-layer optimization framework;

[0106] 6) Alarm and Log Module: Records equipment anomalies and scheduling events, displays operating status via LED indicators, and stores logs for ≥30 days.

[0107] Implementation method:

[0108] 1. Device Hardware Deployment

[0109] 1) The core control module adopts a high-performance core board, configured with 1GB RAM and 8GB ROM, and the IO pin voltage is uniformly 3.3V, reducing the complexity of device adaptation;

[0110] 2) The metrology module uses a hard metrology chip, the built-in CT magnetic core meets the corresponding working temperature requirements, the accuracy class is ≥0.2, and the external CT is selected as 250A / 5A specification, which complements the built-in CT;

[0111] 3) The communication module is equipped with an external WiFi antenna, which is led out of the power distribution box to ensure signal strength; the 4G module is expanded through the MiniPCIe interface and supports NBIoT / 4G dual-mode switching;

[0112] 4) The interface module uses nickel-plated copper terminals. The high-voltage terminal hole size is 8.5*8.5mm, the depth is 14mm, and it supports 35mm... 2 The wires are securely connected;

[0113] 5) The device casing is made of PC + 30% glass fiber material, with a flame retardant rating of UL94V-0. It has passed salt spray and vibration tests and is suitable for the installation environment of household distribution boxes.

[0114] 2. Scheduling method execution flow

[0115] Taking the whole-house backup power + VPP response scenario as an example, the specific execution steps are as follows:

[0116] 1) Data acquisition phase: The metering module collects data such as grid voltage, photovoltaic output power, energy storage SOC, charging pile status, and heat pump power consumption every 100 milliseconds. After digital filtering, the data is uploaded to the core control module. At the same time, the module obtains the daily dynamic electricity price curve and VPP dispatch instructions (such as 2kWh of electricity needs to be discharged from 14:00 to 15:00) from the cloud platform via Ethernet.

[0117] 2) Outer layer optimization stage: The GA-LP algorithm generates a daily scheduling plan based on dynamic electricity price, VPP instructions and equipment parameters: During off-peak hours (00:00-08:00), the energy storage is controlled to charge to 90% SOC, and the charging pile starts charging from 06:00-07:00; During peak hours (10:00-14:00, 19:00-22:00), the energy storage is controlled to discharge, and the heat pump stops to avoid peak hours from 13:00-14:00.

[0118] 3) Inner layer control stage: At 14:00, the MPC module detected that the photovoltaic output power was 1.8kW and the household load was 1.2kW, with a risk of 0.6kW reverse current. It immediately issued a command to limit the inverter output power to 1.2kW, with a response time of 1.8 seconds. At the same time, according to the VPP discharge command, it controlled the energy storage to release 0.8kW of power to ensure that the total discharge meets the requirements.

[0119] 4) Abnormal response phase: At 14:30, the power grid suddenly loses power. The voltage detection circuit identifies the power outage state within 150 milliseconds. The core control module immediately issues an instruction to disconnect the power grid connection. At the same time, according to the load priority sequence, low-priority loads (such as smart home appliances and general lighting) are disconnected, and only critical loads such as refrigerators and routers are retained. Energy storage and diesel generator work together to supply power. The switching completion time is 180 milliseconds.

[0120] 5) Adaptive adjustment phase: By analyzing users' electricity consumption behavior through reinforcement learning algorithms, it was found that users habitually use the dishwasher between 18:00 and 19:00. The load priority during this period is then automatically adjusted to avoid unnecessary downtime.

[0121] 3. Test Verification Results

[0122] 1) Metering accuracy test: Under the reference conditions of 220V, 5A, and power factor 1, the active power error is ±0.3% and the reactive power error is ±1.5%, which meets the requirements of Level 1 active power accuracy and Level 2 reactive power accuracy.

[0123] 2) Dispatch optimization test: Under dynamic electricity pricing scenarios, energy costs are reduced by 25%-30% compared to traditional solutions; under VPP response scenarios, dispatchable capacity is increased by 55%, and command response completion rate is 100%.

[0124] 3) Response speed test: The average response time of the anti-reverse current control is 1.6 seconds, and the average time of islanded backup power switching is 170 milliseconds, which meets the design requirements;

[0125] 4) Communication stability test: WiFi communication distance is 3 meters when the power distribution box is closed, RS485 communication distance is 120 meters, and there is no packet loss in continuous data transmission for 72 hours;

[0126] 5) Reliability test: After 500 hours of continuous operation in an environment with a high temperature of 85℃ and a high humidity of 85%, there were no functional abnormalities or hardware failures.

[0127] 6) Communication Reliability Test: Under a standard laboratory environment (simulating the shielding of a home's walls and electrical distribution box), a 72-hour continuous large data packet transmission stress test was conducted on a 2.4GHz WiFi network. The test results show that, with the electrical distribution box closed, the average data transmission success rate between the device and the external antenna access point reaches 99.95%, and the effective communication distance remains stable at over 3 meters, meeting the reliable communication requirements in complex home environments.

[0128] It should be understood that those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A dual-closed-loop optimal scheduling method for household energy, characterized in that, Includes the following steps: S1: Multi-source data acquisition and preprocessing: Real-time operation data of power grid, photovoltaic, energy storage and household electrical equipment are acquired through the metering module, while dynamic electricity price data and virtual power plant (VPP) dispatch instructions are obtained. S2: Outer Global Optimization Scheduling: Based on a hybrid optimization model of genetic algorithm and linear programming, with the goal of minimizing energy costs and maximizing VPP response benefits, it generates future equipment operation plans by taking into account dynamic electricity price curves, VPP scheduling instructions, equipment parameters and user needs. S3: Inner Real-Time Control Closed Loop: Employs model predictive control technology to dynamically correct the outer plan based on real-time data, and executes anti-backflow control, islanded backup power control, and VPP command response; S4: Status Feedback and Adaptive Adjustment: Collect device status and command execution results, update and optimize model parameters through reinforcement learning algorithms, and optimize load priority weights based on user electricity consumption behavior data.

2. The dual closed-loop optimization scheduling method for household energy according to claim 1, characterized in that: In step S1, the real-time operating data includes the controller's own power consumption; The collected data undergoes digital filtering, harmonic correction, and outlier removal. Dynamic electricity price data includes day-ahead spot electricity prices and intraday real-time price correction data at 15-minute intervals.

3. The dual closed-loop optimization scheduling method for household energy according to claim 1, characterized in that: In step S2, the equipment operation plan includes the charging and discharging periods and power of the energy storage battery, the start-up time of the electric vehicle charging pile, and the start-up and shutdown periods of controllable loads including heat pumps and smart home appliances.

4. The dual closed-loop optimization scheduling method for household energy according to claim 1, characterized in that: In step S3, the anti-reverse flow control controls the inverter output power through a proportional-integral-derivative PID regulation algorithm; The islanded backup power control disconnects low-priority load circuits in sequence according to the energy storage state of charge (SOC) and the preset load priority sequence to ensure power supply to critical loads. VPP command response decomposes scheduling commands into energy storage charging and discharging power and controllable load adjustment, and dynamically allocates adjustment resources.

5. The dual closed-loop optimization scheduling method for household energy according to claim 1, characterized in that: Data transmission is achieved through a multi-mode communication network that supports Ethernet, WiFi, 4G, RS485, LAN, and BPLC power line carrier. The WiFi uses an external antenna to ensure a communication distance of no less than 2 meters when the distribution box is closed.

6. An intelligent power control device for implementing the dual closed-loop optimization scheduling method for household energy as described in any one of claims 1-5, characterized in that, include: The core control module is used to run the dual closed-loop optimization scheduling algorithm; The metering module is used to collect end-to-end power data, including the device's own power consumption. A multimodal communication module is used for data interaction with the power grid platform, VPP platform, and household electrical appliances; Interface and structure modules are used to provide device access interfaces and meet physical installation and environmental protection requirements.

7. The intelligent power control device according to claim 6, characterized in that: The metering module is equipped with a hard metering chip and supports 1P2W, 3P3W or 3P4W wiring methods. Its active power metering accuracy is not lower than Class 1 and its reactive power metering accuracy is not lower than Class 2. The metering current transformer (CT) is located upstream of the device's power supply node and includes both built-in and external CTs; the module is also equipped with a hardware clock with temperature compensation function.

8. The intelligent power control device according to claim 6, characterized in that: The multimodal communication module includes: The northbound communication unit supports Ethernet, external antenna WiFi, and optional 4G communication, and is compatible with MQTT, IEC-61870-5-104, and OpenADR protocols. The southbound communication unit includes multiple RS485 interfaces, a LAN port, and an optional BPLC carrier interface, supporting Modbus-RTU / TCP protocols; The near-end communication unit adopts WiFi access point (AP) mode for local parameter configuration.

9. The intelligent power control device according to claim 6, characterized in that: The interface and structural modules are mounted on DIN35 rails, the housing width is no more than 108 mm, and the protection level is IP30. Provides digital input / output (DI / DO), 12-volt DC input / output, and pulse output interfaces, supporting wires with a cross-sectional area of ​​not less than 35 square millimeters; The outer shell is made of flame-retardant polycarbonate (PC) with added glass fiber, and the operating temperature range is -25 degrees Celsius to 70 degrees Celsius.

10. The intelligent power control device according to claim 6, characterized in that: It also includes software functional modules, which at least include: The module includes a data acquisition and preprocessing module, a dual-closed-loop scheduling module integrating an outer layer genetic algorithm-linear programming (GA-LP) optimization and an inner layer model predictive control (MPC) algorithm, an equipment management module for managing inverters, charging piles and controllable loads, a VPP coordination module for parsing and responding to VPP commands, a dynamic electricity price management module for managing dynamic electricity price data, and an alarm and log module for recording operating status and anomalies.

Images