Energy efficiency dual-carbon management system and method

Abstract

The invention discloses an energy efficiency dual-carbon management system and method, and the method comprises the steps: predicting a building load demand through a time sequence prediction model, and generating and executing a dynamic regulation and control instruction of an air conditioner and illumination equipment in combination with a reinforcement learning algorithm; according to the actual energy consumption of the equipment, calculating regional carbon emission, generating a hash value from carbon emission data and associated data, writing the hash value into a block chain, and verifying the data consistency through an intelligent contract; and based on the load prediction result, the time-of-use electricity price and the renewable energy output prediction, constructing a multi-target optimization model to solve the energy storage charging and discharging power, the renewable energy output power and the power grid electricity purchasing power, and generating a collaborative scheduling instruction. The air conditioner temperature adjustment amount and the lighting brightness level are generated in real time through a reinforcement learning algorithm, compared with traditional PID control, energy consumption is lower, and meanwhile the sensible temperature of a user is kept within a comfortable interval; and a multi-objective optimization model containing low carbon, economy and energy storage health degree is constructed, and the output proportion of photovoltaic, energy storage and a power grid is dynamically coordinated.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of double carbon management, and particularly relates to an energy efficiency double carbon management system and method. BACKGROUND

[0002] Double carbon management generally refers to carbon emission and carbon neutralization management, that is, reducing greenhouse gas emissions and taking measures to offset the emissions that have been generated. With the development of economy, green and low-carbon development has become one of the important directions in economic development, and energy-saving electricity as a key direction of carbon emission reduction has a profound impact on achieving the goal of green and low-carbon development.

[0003] In the prior art, building energy efficiency optimization relies on single device upgrade or static threshold control, such as fixed temperature setting, preset lighting mode, etc., and it is difficult to adapt to dynamic environmental changes and personnel flow, resulting in limited energy efficiency improvement. For example, the traditional central air conditioning system often runs at a fixed power, and cannot adjust the refrigeration load according to the real-time flow density and outdoor temperature and humidity, causing energy waste; the lighting system generally lacks coordinated control with natural light and personnel flow distribution, and the phenomenon of excessive lighting is prominent. In addition, carbon emission accounting mostly uses offline statistics or theoretical estimation, and the data is isolated and easy to be tampered with. In terms of energy dispatching, the existing scheme mostly focuses on a single target (such as economy), lacks comprehensive optimization of low carbon and device health, resulting in problems such as frequent overcharging / overdischarging of energy storage systems and low photovoltaic consumption rate. Therefore, an intelligent management scheme integrating dynamic sensing, reliable traceability and multi-target coordination is urgently needed to break through the data island, improve energy efficiency and ensure carbon emission compliance. SUMMARY

[0004] The purpose of the present application is to provide an energy efficiency double carbon management system and method to solve the problems raised in the background art.

[0005] To achieve the above-mentioned purpose, the present application provides the following technical scheme: an energy efficiency double carbon management method, comprising the following steps: S1, collecting real-time environmental parameters, device operating states and personnel density data of each area in the building through a multi-source data acquisition system; S2, based on the collected data, using a time series prediction model to predict the building load demand, combining a reinforcement learning algorithm to generate dynamic control instructions for air conditioning and lighting devices and executing the instructions; S3, calculating the regional carbon emissions according to the actual energy consumption of the devices, generating a hash value of the carbon emission data and associated data and writing the hash value into a blockchain, and verifying the data consistency through a smart contract; S4, based on the load prediction results, time-of-use electricity prices and renewable energy output prediction, constructing a multi-objective optimization model to solve the energy storage charging and discharging power, renewable energy output power and grid purchase power, and generating a coordinated dispatching instruction; S5, feedback the real-time carbon emission to step S2 to dynamically adjust the carbon emission weight coefficient, and update the carbon emission factor model according to the actual data of the power grid.

[0006] Preferably, in step S1, the environmental parameters at least include indoor and outdoor temperature and humidity, CO2 concentration; the equipment operating state at least includes air conditioning power and lighting power; and the personnel density data is obtained through a personnel activity monitoring device.

[0007] Preferably, step S2 comprises: normalizing and processing missing values of the collected data; using an LSTM network containing a CO2 concentration correction mechanism to predict building load demand; taking the weighted sum of energy efficiency rewards and carbon emission penalties as an objective function, and generating device control instructions through reinforcement learning.

[0008] Preferably, in step S3, the carbon emission calculation satisfies: ; Wherein, is the personnel density data, and the actual energy consumption of the equipment , is the carbon emission factor, and η is the additional carbon emission coefficient per capita.

[0009] Preferably, in step S3, the blockchain verification comprises: when the smart contract detects that the on-chain hash value is inconsistent with the actual data, triggering an early warning and suspending the control strategy.

[0010] Preferably, the objective function of the multi-objective optimization model of step S4 is: ; Wherein, α, β, δ are weight coefficients, is the power grid purchase power, is the state of charge of the energy storage, is the ideal value of the state of charge of the energy storage.

[0011] An energy efficiency double carbon management system comprises: A multi-source data acquisition module is used to acquire environmental parameters, equipment operating states and personnel density data in each area of a building in real time through a multi-source data acquisition system; A dynamic optimization control module is used to predict building load demand by using a time series prediction model, generate dynamic control instructions for air conditioning and lighting equipment by using a reinforcement learning algorithm, and execute the instructions; A carbon emission tracing module is used to calculate regional carbon emissions according to the actual energy consumption of the equipment, write the carbon emission data and associated data into a blockchain by generating a hash value, and verify the data consistency by using a smart contract. An energy coordination scheduling module is configured to construct a multi-objective optimization model to solve the energy storage charging and discharging power, the renewable energy output power and the grid power purchase power based on the load prediction result, the time-of-use electricity price and the renewable energy output prediction, and generate a coordination scheduling instruction. A closed-loop feedback and strategy correction module is configured to feed back the real-time carbon emission to the dynamic optimization control module to dynamically adjust the carbon emission weight coefficient, and update the carbon emission factor model according to the actual data of the power grid.

[0012] The technical effects and advantages of the present application are as follows: 1. The air conditioner temperature adjustment amount and lighting brightness level are generated in real time by the reinforcement learning algorithm, and the equipment operation is dynamically controlled in combination with the environmental parameters (temperature and humidity, CO2 concentration) and the personnel density, which is lower in energy consumption compared with the traditional PID control, and the user's body temperature is maintained in the comfortable interval; The sensor data hash value and the carbon emission calculation result are stored by using the blockchain technology, so that the data cannot be tampered with, and the data consistency is automatically verified by using the smart contract, so that false reporting or errors are avoided. A multi-objective optimization model containing low carbon, economy and energy storage health is constructed to dynamically coordinate the output proportion of photovoltaic, energy storage and power grid. BRIEF DESCRIPTION OF DRAWINGS

[0013] Figure 1 The method flowchart of the present application is provided. DETAILED DESCRIPTION

[0014] The specific embodiments of the present application will be further described below in combination with the drawings. It should be noted that the description of these embodiments is used to help understand the present application, but does not constitute a limitation on the present application. In addition, the technical features involved in each embodiment of the present application described below can be combined with each other as long as they do not conflict with each other.

[0015] The present application provides an energy efficiency double carbon management method as shown in Figure 1 , which comprises the following steps: Step 1: Real-time acquisition of environmental parameters (including indoor temperature , outdoor temperature , indoor humidity , outdoor humidity , CO2 concentration C), equipment operating state (including air conditioner power , lighting power ) and personnel density data is achieved by integrating a multi-source data acquisition module of temperature and humidity sensor, infrared thermal imager, CO2 concentration sensor and people recognition equipment. Step 2: Receiving the output data of the multi-source data acquisition module, predicting the building load demand , and sends the control instructions to the building equipment execution terminal; Specifically, the LSTM neural network is used to predict the building load demand The specific steps of the prediction model are as follows: The data collected by the input multi-source data acquisition module; The input data is normalized, and the short-time missing data of the sensor is filled by linear interpolation, and the long-time missing period is marked as abnormal and excluded; The LSTM network structure of the input layer, hidden layer and output layer is constructed; Based on the above-mentioned LSTM network structure, a CO2 concentration correction mechanism is introduced to obtain a prediction model for predicting building load demand based on LSTM neural network:

[0016] t is the time variable, and epsilon is a trainable parameter that learns the nonlinear effect of CO2 concentration on the intensity of human activity through backpropagation; The reinforcement learning algorithm is used to generate device control instructions, including air conditioner set temperature adjustment And the lighting brightness level The objective function is: ; Where is the energy saving reward, is the carbon emission penalty term, and gamma is the weight coefficient; Dynamic balance of building energy efficiency and user comfort, while providing a decision basis for subsequent carbon emission optimization and energy scheduling.

[0017] Step 3, according to the actual energy consumption of the equipment , calculate the regional carbon emissions:

[0018] Where, is the carbon emission factor, eta is the additional carbon emission coefficient per capita, is the reference personnel density; Carbon emission data and associated sensor data are used to generate hash values using the SHA-256 algorithm and written to the blockchain, triggering the smart contract to verify data consistency; the smart contract is an automatically executed contract clause that automatically triggers the corresponding operation when the preset conditions are met, without the need for third-party intervention; ensure that the data written to the blockchain is consistent with the actual collected data, prevent tampering or errors. If the data is inconsistent, trigger the early warning mechanism to notify the administrator to check the sensor or network problem, and suspend the related control strategy until the data is restored to be consistent.

[0019] Through blockchain storage and smart contract verification, the system realizes the non-tamperable, transparent and traceable of carbon emission data; the hash value ensures data integrity, and the smart contract automatic verification improves efficiency. This mechanism is deeply coupled with dynamic optimization control and energy scheduling module, which guarantees the efficient and low-carbon operation of the whole system on the basis of data credibility, while meeting the strict regulatory compliance requirements.

[0020] It should be noted that during the actual energy consumption calculation process of the equipment, if the power value of a certain equipment exceeds the rated value ± 20%, an alarm is triggered and the abnormal period data is excluded; if there is a short-time data missing situation, the adjacent period mean interpolation processing is adopted; if there is a long-time data missing situation, the device state is checked by artificial.

[0021] Step 4, based on the load prediction result , time-of-use electricity price and photovoltaic output prediction model , wherein GHI is the global horizontal irradiance; According to the above parameters, a multi-objective optimization problem is constructed: , wherein α, β, δ are weight coefficients, is the grid purchase power, is the state of charge of the energy storage, ideal value of the state of charge of the energy storage; Solving the multi-objective optimization problem obtains the energy storage charging and discharging power, the photovoltaic inverter output power and the grid purchase power, generates the cooperative scheduling instruction, and performs scheduling adjustment of the energy storage charging and discharging power, the photovoltaic inverter output power and the grid purchase power; Step 5, The real-time carbon emission calculated in step 3 is fed back to step 2 to dynamically adjust the carbon emission weight γ in the reinforcement learning objective function; According to the grid carbon emission factor corrected according to the actual operation data of the energy storage, the carbon emission model of step 3 is updated.

[0022] The present application also provides an energy efficiency double carbon management system, comprising: Multi-source data acquisition module: integrating temperature and humidity sensor, infrared thermal imager, carbon dioxide concentration sensor and face recognition device, real-time acquisition of building internal regional environment parameters, equipment operation state and personnel density data; Dynamic optimization control module: based on LSTM neural network to predict building load demand, combined with reinforcement learning algorithm to generate dynamic control instructions of air conditioning and lighting equipment; Carbon emission traceability module: using blockchain technology to store energy consumption data and carbon emission factor on the chain, realizing the whole life cycle traceability of carbon footprint; Energy co-scheduling module: according to the time-of-use electricity price and renewable energy output prediction, optimize the charging and discharging strategy of energy storage system and the power output of photovoltaic inverter.

[0023] Closed-loop feedback and strategy correction module: real-time data update for carbon emission model update.

[0024] Although the embodiments of the present application have been shown and described above, it should be understood by those skilled in the art that the above embodiments are exemplary and cannot be construed as limiting the present application, and those skilled in the art can make changes, modifications, replacements and variations to the above embodiments within the scope of the present application.

Claims

1. An energy efficiency and dual carbon management method, characterized in that: The following steps are involved: S1. Real-time collection of environmental parameters, equipment operating status, and personnel density data for each area within the building through a multi-source data acquisition system; S2. Based on the collected data, the time series prediction model is used to predict the building load demand, and the reinforcement learning algorithm is used to generate and execute dynamic control instructions for air conditioning and lighting equipment; S3. Calculate regional carbon emissions based on the actual energy consumption of the equipment, generate hash values ​​for the carbon emission data and related data, and write them into the blockchain. Verify data consistency through smart contracts. S4. Based on load forecast results, time-of-use electricity prices, and renewable energy output forecasts, a multi-objective optimization model is constructed to solve the energy storage charging and discharging power, renewable energy output power, and grid power purchase power, and generate coordinated dispatch instructions; S5. Feedback the real-time carbon emissions to step S2 to dynamically adjust the carbon emission weight coefficient, and update the carbon emission factor model based on the actual data of the power grid.

2. The energy efficiency and carbon management method according to claim 1, characterized in that: In step S1: the environmental parameters include at least indoor and outdoor temperature and humidity, and CO2 concentration; the equipment operating status includes at least air conditioning power and lighting power; and the personnel density data is obtained through personnel activity monitoring equipment.

3. The energy efficiency and carbon management method according to claim 1, characterized in that: The step S2 includes: normalizing the collected data and processing missing values; using an LSTM network including a CO2 concentration correction mechanism to predict building load demand; and generating equipment control instructions through reinforcement learning using the weighted sum of energy efficiency rewards and carbon emission penalties as the objective function.

4. The energy efficiency and carbon management method according to claim 1, characterized in that: The carbon emissions calculation in step S3 satisfies: ； in, is the personnel density data, the actual energy consumption of the equipment , is the carbon emission factor, and η is the per capita additional carbon emission coefficient.

5. The energy efficiency and carbon management method according to claim 1, characterized in that: The blockchain verification in step S3 includes: when the smart contract detects that the hash value on the chain is inconsistent with the actual data, triggering an early warning and suspending the control strategy.

6. The energy efficiency and carbon management method according to claim 1, characterized in that: The objective function of the multi-objective optimization model in step S4 is: ； Among them, α, β, and δ are weight coefficients, Purchase power for the grid, is the energy storage charge state, Ideal value of energy storage state of charge.

7. An energy efficiency and dual carbon management system, characterized in that: include: Multi-source data acquisition module: used to collect environmental parameters, equipment operating status and personnel density data of various areas in the building in real time through the multi-source data acquisition system; Dynamic Optimization Control Module: This module uses a time series prediction model to predict building load demand and combines it with a reinforcement learning algorithm to generate and execute dynamic control instructions for air conditioning and lighting equipment. Carbon emission traceability module: used to calculate regional carbon emissions based on the actual energy consumption of the equipment, generate hash values ​​for carbon emission data and related data, and write them into the blockchain, verifying data consistency through smart contracts; Energy collaborative dispatch module: This module is used to build a multi-objective optimization model based on load forecast results, time-of-use electricity prices, and renewable energy output forecasts to solve the energy storage charging and discharging power, renewable energy output power, and grid power purchase power, and generate collaborative dispatch instructions. Closed-loop feedback and strategy correction module: Feedback real-time carbon emissions to the dynamic optimization control module to dynamically adjust the carbon emission weight coefficient, and update the carbon emission factor model based on the actual data of the power grid.

Images