An energy management system for industrial AI control

Abstract

The application discloses an energy management system for industrial AI control and relates to the technical field of energy management, comprising a state acquisition and executable control range generation module, which is used for accessing the real-time running states of an industrial energy management system, a supervisory control and data acquisition system, energy storage management equipment and edge control equipment, and generating an executable control range in a current control period, which allows an artificial intelligence model to participate in selection according to the real-time running states; a strategy preference generation module, which is used for receiving user operation targets and the executable control range, and making the artificial intelligence model generate strategy preference information in the executable control range; according to the application, the industrial energy management system generates an executable control range according to real-time states on site, the artificial intelligence model only outputs strategy preference information in the executable control range, and then a safe authorization side reversely generates device-level control instructions and performs short-time effective execution.

Description

Technical Field

[0001] This invention relates to the field of energy management technology, specifically to an energy management system for industrial AI control. Background Technology

[0002] With the application of artificial intelligence (AI) technology in microgrids, virtual power plants, energy storage power stations, zero-carbon parks, and industrial energy management scenarios, energy management systems are gradually incorporating large language models, reinforcement learning models, and predictive optimization models to assist in generating energy storage charging and discharging plans, multi-resource collaborative scheduling strategies, energy consumption plan correction strategies, and operational optimization suggestions. Industrial sites typically deploy energy management systems, monitoring and data acquisition systems, energy storage battery management systems, power conversion devices, and edge control devices to collect on-site operating status, execute scheduling plans, and maintain equipment operational safety. The combination of AI strategy generation capabilities and the stable execution capabilities of industrial control systems can provide new technical pathways for optimizing the operation of complex energy systems.

[0003] In industrial energy control scenarios, if AI models directly generate device-level control commands for energy storage, loads, photovoltaic systems, charging piles, and other adjustable energy devices, inconsistencies can easily arise between the command content and the real-time status of the equipment, grid operating procedures, on-site communication status, and safety margins. If rule verification and interception logic are only set after the AI ​​commands are generated, it remains a post-hoc approach, failing to prevent the AI ​​model from generating out-of-bounds control intentions at the source. Furthermore, when the on-site status changes rapidly, verified control commands may lose their basis for safe execution due to status expiration. Existing energy management systems also face the challenge of non-intrusive integration of AI strategies into existing control links. It is necessary to preserve the security boundaries of the original industrial control system while allowing AI to participate in operational optimization. Therefore, an industrial AI energy control architecture is needed that allows AI to participate in strategy selection without directly acquiring device control. Summary of the Invention

[0004] The purpose of this invention is to provide an energy management system for industrial AI control to solve the problems mentioned in the background art.

[0005] To address the aforementioned technical problems, this invention provides the following technical solution: an energy management system for industrial AI control, comprising: a status acquisition and executable control range generation module, used to access the real-time operating status of an industrial energy management system, a monitoring and data acquisition system, an energy storage management device, and an edge control device, and generate an executable control range that allows the artificial intelligence model to participate in selection within the current control cycle based on the real-time operating status; a strategy preference generation module, used to receive the user's operating objective and the executable control range, enabling the artificial intelligence model to generate strategy preference information within the executable control range, wherein the strategy preference information represents the selection order and recommendation reasons for different executable control segments; and an authorized execution and feedback correction module, used to generate device-level control instructions from the executable control range based on the strategy preference information, receive feedback results after device execution, and update the executable control range for the next control cycle based on the feedback results.

[0006] According to the above technical solution, the status acquisition and executable control range generation module includes a status access unit, a boundary conversion unit, and a control segment generation unit. The status access unit is used to acquire the current output, equipment operating limits, equipment charge status, communication status, and alarm status of the adjustable energy equipment. The boundary conversion unit is used to convert the equipment operating limits and on-site operating status into the power range that can be adjusted within the current control cycle. The control segment generation unit is used to combine the equipment object, the allowable adjustment range, the effective period, and the failure condition into an executable control segment. The strategy preference generation module includes a requirement parsing unit, a fragment sorting unit, and an output type restriction unit. The requirement parsing unit is used to convert the user-inputted running objective into an energy control objective. The fragment sorting unit is used to enable the artificial intelligence model to sort the executable control fragments according to the energy control objective. The output type restriction unit is used to restrict the artificial intelligence model to output only strategy preference information and not device-level control instructions. The authorized execution and feedback correction module includes an instruction synthesis unit, an edge distribution unit, an execution feedback unit, a failure rollback unit, and a process recording unit. The instruction synthesis unit is used to convert the authorized executable control segments into device-level control instructions. The edge distribution unit is used to distribute the device-level control instructions to field devices. The execution feedback unit is used to collect field feedback results after execution. The failure rollback unit is used to switch to a preset safety control state when the executable control segment fails. The process recording unit is used to record the executable control range, strategy preference information, device-level control instructions, and feedback results.

[0007] According to the above technical solution, the system performs the following steps: S1. Generate executable control range: The status acquisition and executable control range generation module acquires the real-time operating status of adjustable energy equipment on site and reads the equipment operating limits and safety control rules configured in the industrial energy management system. The system determines the direction and range that each device can be safely adjusted within the current control cycle based on the current equipment status. The system forms executable control segments by including the equipment objects that can be safely adjusted, the allowed adjustment content, the effective period, and the failure conditions, and summarizes them into executable control ranges. S2. Generate strategy preference information: The strategy preference generation module receives the user's running goal and the executable control range, and converts the user's running goal into a strategy selection task that the artificial intelligence model can understand. The artificial intelligence model must not directly output device-level control instructions, but can only sort and describe the executable control segments within the executable control range. The system thus obtains strategy preference information, which indicates which executable control segments are preferentially selected to approach the user's running goal. S3. Generate and issue device-level control commands: After receiving the policy preference information, the authorization execution and feedback correction module reads the executable control segments in sequence according to the sorting results. The system only authorizes the executable control segments that are still in effect and have not triggered the failure condition, and converts the authorized executable control segments into device-level control commands that the field devices can recognize. The device-level control commands are sent to the corresponding field devices by the edge issuing unit. The artificial intelligence model does not participate in the direct issuance of the final command. S4. Collect execution feedback and handle failure situations: The execution feedback unit collects the output feedback, status feedback, communication feedback and alarm feedback after the field equipment is executed. The system will make a consistency judgment between the feedback results and the issued equipment-level control commands to confirm whether the equipment execution results are still in a safe control state. If the executable control segment fails, the failure rollback unit will revoke the subsequent authorization and switch the corresponding equipment to the preset safe control state. S5. Update the executable control range for the next control cycle: The system regenerates the executable control range for the next control cycle based on the execution feedback results, the current status of the equipment, and the field control rules. The process recording unit records the input status, executable control range, strategy preference information, equipment-level control instructions, and feedback results for each control cycle. The system thus forms a closed-loop control process from status acquisition, range generation, preference selection, authorized execution to feedback update.

[0008] According to the above technical solution, step S1 includes: S1-1. Collect the equipment operating status of the current control cycle: The system will... An adjustable energy device in the current control cycle The current output is recorded as , the first The lower limit of the power output of each adjustable energy device is denoted as: The upper limit of device power is denoted as The upward adjustment of the climbing limit is recorded as... The positive value of adjusting the downward climbing limit is recorded as . The duration of the control cycle is recorded as The lower limit of the permissible power under the condition, calculated based on the equipment's state of charge, voltage, current, and temperature protection conditions, is denoted as... The maximum permitted power for the state is denoted as The lower limit and upper limit of the state-permitted power are used to uniformly convert the state of charge, voltage, current and temperature protection conditions into executable boundaries in the form of power, so that subsequent calculations can determine whether the device has safe adjustment capabilities within the same power dimension. The ramp limit represents the maximum change capability of the adjustable energy device to increase or decrease the output power within a control cycle. The state of charge is used to represent the proportion of the current remaining power of the energy storage device to the available capacity. S1-2. When calculating the allowable adjustment range within the current control cycle, the system does not directly use the upper and lower limits of the equipment's rated power as the selection range for the artificial intelligence model. Instead, it simultaneously considers the equipment power boundary, ramp-up capability boundary, and field condition permissible boundary, and selects the adjustment range that is satisfied by all three types of boundaries. Based on this processing, the system calculates the minimum allowable adjustment amount of the i-th adjustable energy device within the current control cycle t using the following formula. The system calculates the first [item] based on the same device boundary conditions. An adjustable energy device in the current control cycle Maximum allowable adjustment within The significance of the above two formulas is that they unify the power boundary of the equipment itself, the ramp capability boundary, and the field condition conversion boundary into the power adjustment amount that can be increased and decreased within the current control cycle. This allows the subsequent artificial intelligence model to read only the adjustment range that has met the field safety conditions. Through the above maximum and minimum value calculations, the system compresses multiple safety conditions into a continuous allowable adjustment range. Only the adjustment amount that falls into this range is allowed to enter the subsequent strategy preference ranking process. S1-3, Generate executable control segment: When Not greater than When the system indicates that the i-th adjustable energy device has a safe adjustment range within the current control cycle, the system encapsulates the i-th adjustable energy device, the minimum allowable adjustment amount, the maximum allowable adjustment amount, the current control cycle, and the set of failure conditions into an executable control segment. ,in, Indicates the first An adjustable energy device in the current control cycle The control segments that can be authorized for use within. This indicates the failure judgment conditions that the executable control segment needs to continuously meet within the current control cycle, including at least the execution deviation exceeding the limit condition, the communication delay exceeding the limit condition, and the field alarm trigger condition. The significance of this formula is that it binds the safety adjustment range and the failure judgment conditions together, so that the artificial intelligence model can only select control segments with time constraints and failure constraints, and cannot rewrite the equipment adjustment amount alone.

[0009] According to the above technical solution, step S2 includes: S2-1. Converting User Operating Objectives into Planned Deviation Correction Objectives: To avoid the artificial intelligence model directly generating device-level control commands based on natural language objectives, the system first converts the user operating objectives into planned deviation correction objectives that the industrial energy management system can quantify. These planned deviation correction objectives represent the degree of deviation between the current total output and the planned output, and serve as a unified basis for the subsequent sequencing of executable control segments. The system will then use the current control cycle... The planned output is recorded as The total output measured on site will be recorded as The planned deviation for the current control cycle is calculated using the following formula. ,in, This formula represents the output deviation that needs to be corrected within the current control cycle. The significance of this formula is to convert the user-proposed economic operation, plan tracking, and scheduling correction targets into power deviation correction amounts that the field control system can recognize. Through this calculation, the abstract operation target input by the user is converted into a specific power deviation amount, so that the artificial intelligence model can only sort segments around the deviation correction effect instead of directly generating control commands for the field equipment. S2-2. Calculate the correction contribution of each executable control segment to the plan deviation: The system generates candidate adjustment amounts within the allowable adjustment interval corresponding to each executable control segment Cᵢ(t). The candidate adjustment values, along with their corresponding executable control segments, are then provided to the artificial intelligence model for sorting; the artificial intelligence model does not change... and Nor will it generate new equipment adjustments outside the permitted adjustment range. The system calculates the correction contribution of the candidate adjustments to the plan deviation according to the following formula. ,in, This indicates the reduction in planning deviation after adopting the candidate adjustment amount. The significance of this formula is that it allows the ranking criteria of the artificial intelligence model to directly correspond to the planning deviation correction target in the industrial field, rather than outputting equipment commands that deviate from the field control boundary. S2-3, Generate policy preference information: The system will calculate the correction contributions of each executable control segment. and the corresponding executable control segment Input the AI ​​model and make it output a sequence of policy preferences. ,in, This indicates the ranking of executable control segments recommended by the artificial intelligence model within the current control cycle. to This represents the sorted adjustable energy device numbers. The significance of this formula lies in restricting the output of the artificial intelligence model to the sorting result, allowing the AI ​​model to express policy preferences rather than directly generating device-level control commands. The policy preference sequence... It only indicates the order of use of executable control segments recommended by the artificial intelligence model, and does not include the target output of the equipment, the switching actions of the equipment, or the execution format that can be directly sent to the field equipment.

[0010] According to the above technical solution, step S3 includes: S3-1. Selecting the authorized control segment based on the policy preference sequence: The authorization execution and feedback correction module receives the policy preference sequence. Subsequently, instead of directly executing the policy preference sequence, the system reads executable control segments one by one according to the order of the policy preference sequence. For each executable control segment read, the system first confirms that the segment is still within the current control cycle and its failure condition set. Not triggered; only executable control segments that meet the above conditions are marked as authorized control segments. ,in, Indicates the first The formula defines a control segment that is safely authorized for use by an adjustable energy device within the current control cycle. The significance of this formula is that it clarifies that the authorized control segment can only come from the executable control range that has been generated in step S1, and the artificial intelligence model cannot add control content outside of this range. S3-2, Generating device-level control commands from authorized control segments: In authorized control segments Once determined, the system selects the actual adjustment amount from the candidate adjustment amounts corresponding to the authorized control segment. and ensure Still falling into to Within the defined range, based on the actual adjustment amount, the system generates a target output that the field equipment can recognize. ,in, Indicates from the authorized control segment The actual adjustment amount selected in the middle, Indicates that it is sent to the first The target output of an adjustable energy device is defined by this formula, which transforms the strategy preferences recommended by the artificial intelligence model into a target output that the industrial field device can execute, while ensuring that the target output comes from the already authorized adjustment amount. It is not a control result directly issued by the artificial intelligence model, but a device execution target calculated by the security authorization side based on the authorized control segment; S3-3, Generate and issue device-level control commands: The system will... Equipment number and target output of each adjustable energy device Current control cycle and authorized control segment Together, generate device-level control commands ,in, This indicates the device-level control command sent to the field edge control device. The significance of this formula is to bind the authorized source and the target output, so that the field execution device can trace which authorized control segment each command comes from.

[0011] According to the above technical solution, step S4 includes: S4-1, Acquisition of device execution feedback: In device-level control commands After being issued, the execution feedback unit collects the data of the first... The measured output of an adjustable energy device at the next feedback moment The execution deviation is calculated according to the following formula. ,in, Indicates the first The measured output of an adjustable energy device at the next feedback moment. This formula represents the deviation between the actual output of the equipment and the target output. The significance of this formula lies in determining whether the equipment has completed the execution according to the authorized target output. S4-2. Determine if the executable control segment has failed: The system uses the execution deviation exceeding the limit condition and the communication delay exceeding the limit condition as the set of failure conditions. The specific judgment content is used to judge the authorized control segment. Whether there is still a safe basis for continuing execution, the system will record the execution deviation threshold as follows: The feedback communication delay is denoted as The allowed communication delay threshold is denoted as ,when Greater than At that time, the system determines the authorized control segment. Failure; when Greater than At that time, the system determines the authorized control segment. If an authorized control segment is deemed invalid, it will no longer be used as a source of authorization for subsequent control cycles. S4-3, Execution Failure Rollback Control: When at least one of the following conditions is met: execution deviation exceeds the allowable execution deviation threshold, or feedback communication delay exceeds the allowable communication delay threshold, the system determines the authorized control segment. The system has failed and will switch the actual control output at the next feedback moment to the preset safe output; when neither of the above two conditions is met, the system maintains the original target output, and the system will switch the first... The preset safe output of each adjustable energy device is denoted as The actual control output at the next feedback moment is generated according to the following formula. ,in, Indicates the next feedback time for the first... The actual control output adopted by an adjustable energy device is defined by the following formula: when the execution deviation exceeds the allowable execution deviation threshold, the system switches to the safe output preset by the industrial control system; when the communication delay exceeds the allowable communication delay threshold, the system also switches to the safe output preset by the industrial control system; when both the execution deviation and the communication delay are within the allowable range, the system continues to use the authorized target output.

[0012] According to the above technical solution, step S5 includes: S5-1. Update the input state for the next control cycle based on the execution feedback: The system will update the measured output force obtained in step S4. Execution deviation and communication delay Forming a single device feedback status The system then summarizes the individual device feedback states of all adjustable energy devices into the state input set for the next control cycle. ,in This indicates the number of adjustable energy devices currently connected to the system. The significance of this formula is that the executable control range of the next control cycle directly inherits the execution result of the previous control cycle, thus avoiding the artificial intelligence model from continuing to participate in strategy selection based on the expired state. S5-2, Regenerate the executable control range for the next control cycle: The system will set the state inputs. Input the boundary conversion unit from step S1 and recalculate the minimum allowable adjustment for the next control cycle. and maximum allowable adjustment This generates the executable control segment for the next control cycle. This step reauthorizes each control cycle, and the executable control segments from the previous control cycle are not continued by default. S5-3. Record the closed-loop control process: The system summarizes all executable control segments generated within the current control cycle into a control segment set for the current control cycle t. k represents the number of executable control segments generated within the current control cycle, and all device-level control instructions actually issued within the current control cycle are aggregated into the control instruction set for the current control cycle t. This ensures that the process recording object covers the entire control chain from the generation of the executable scope to the issuance of device commands, where m represents the number of device-level control commands actually issued within the current control cycle, forming the control segment set. and the set of control commands Then, the system writes the process record value of the previous control cycle, the state input set of the current control cycle, the control segment set, the strategy preference sequence, the control instruction set, and the state input set of the next control cycle into the hash calculation process to obtain the process record value of the current control cycle t. ,in This represents the process record value from the previous control cycle. The hash calculation function is represented by the above recording method. The system can review afterward whether each device-level control instruction comes from the set of executable control segments generated in the current control cycle, and can confirm whether the artificial intelligence model always outputs the strategy preference sequence only within the set of executable control segments, thereby avoiding the artificial intelligence model from bypassing the executable control range to directly generate device control instructions.

[0013] Compared with existing technologies, the beneficial effects achieved by this invention are as follows: This invention, through a control chain that first generates an executable control range, then receives AI strategy preferences, and finally generates device-level control commands, prevents the AI ​​model from directly outputting executable device commands. This reduces the risk of AI strategies exceeding device power boundaries, voltage boundaries, state of charge boundaries, and scheduling plan boundaries from the source. Since the executable control range is jointly determined by the real-time field status, equipment operating limits, ramp-up capability, communication status, and feedback results, the system can reauthorize executable actions within each control cycle, preventing expired strategies from being issued. This solution enables AI to participate in energy optimization scheduling without large-scale modifications to existing energy management and monitoring systems, while retaining the industrial control system's rights to generate, issue, and roll back final commands, thereby improving the security, interpretability, reproducibility, and engineering implementation stability of industrial AI energy control. Attached Figure Description

[0014] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart illustrating the present invention; Figure 2 This is a schematic diagram of the overall modular structure of the present invention. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] Please see Figure 1 and Figure 2 This invention provides a technical solution: an energy management system for industrial AI control, comprising: a status acquisition and executable control range generation module, used to access the real-time operating status of an industrial energy management system, a monitoring and data acquisition system, an energy storage management device, and an edge control device, and generate an executable control range that allows the artificial intelligence model to participate in selection within the current control cycle based on the real-time operating status; a strategy preference generation module, used to receive the user's operating target and executable control range, enabling the artificial intelligence model to generate strategy preference information within the executable control range, the strategy preference information being used to represent the selection order and recommendation reasons for different executable control segments; and an authorized execution and feedback correction module, used to generate device-level control instructions from the executable control range based on the strategy preference information, receive feedback results after device execution, and update the executable control range for the next control cycle based on the feedback results. The status acquisition and executable control range generation module includes a status access unit, a boundary conversion unit, and a control segment generation unit. The status access unit is used to acquire the current output, equipment operating limits, equipment charge status, communication status, and alarm status of the adjustable energy equipment. The boundary conversion unit is used to convert the equipment operating limits and field operating status into the power range that can be adjusted within the current control cycle. The control segment generation unit is used to combine the equipment object, the allowable adjustment range, the effective period, and the failure condition into an executable control segment. The strategy preference generation module includes a requirement parsing unit, a fragment sorting unit, and an output type restriction unit. The requirement parsing unit is used to convert the user-input operation target into an energy control target. The fragment sorting unit is used to enable the artificial intelligence model to sort the executable control fragments according to the energy control target. The output type restriction unit is used to restrict the artificial intelligence model to output only strategy preference information and not device-level control instructions. The authorized execution and feedback correction module includes an instruction synthesis unit, an edge distribution unit, an execution feedback unit, a failure rollback unit, and a process recording unit. The instruction synthesis unit is used to convert the authorized executable control segments into device-level control instructions. The edge distribution unit is used to distribute the device-level control instructions to the field devices. The execution feedback unit is used to collect the field feedback results after execution. The failure rollback unit is used to switch to a preset safety control state when the executable control segment fails. The process recording unit is used to record the executable control range, strategy preference information, device-level control instructions, and feedback results. The system performs the following steps: S1. Generate executable control range: The status acquisition and executable control range generation module acquires the real-time operating status of adjustable energy equipment on site and reads the equipment operating limits and safety control rules configured in the industrial energy management system. The system determines the direction and range that each device can be safely adjusted within the current control cycle based on the current equipment status. The system forms executable control segments by including the equipment objects that can be safely adjusted, the allowed adjustment content, the effective period, and the failure conditions, and summarizes them into executable control ranges. S2. Generate strategy preference information: The strategy preference generation module receives the user's running goal and the executable control range, and converts the user's running goal into a strategy selection task that the artificial intelligence model can understand. The artificial intelligence model must not directly output device-level control instructions, but can only sort and describe the executable control segments within the executable control range. The system obtains strategy preference information from this, which indicates which executable control segments are preferentially selected to approach the user's running goal. S3. Generate and issue device-level control commands: After receiving the policy preference information, the authorization execution and feedback correction module reads the executable control segments in sequence according to the sorting results. The system only authorizes the executable control segments that are still in effect and have not triggered the failure condition, and converts the authorized executable control segments into device-level control commands that the field devices can recognize. The device-level control commands are sent to the corresponding field devices by the edge issuing unit. The artificial intelligence model does not participate in the direct issuance of the final command. S4. Collect execution feedback and handle failure situations: The execution feedback unit collects the output feedback, status feedback, communication feedback and alarm feedback after the field equipment is executed. The system will make a consistency judgment between the feedback results and the issued equipment-level control commands to confirm whether the equipment execution results are still in a safe control state. If the executable control segment fails, the failure rollback unit will revoke the subsequent authorization and switch the corresponding equipment to the preset safe control state. S5. Update the executable control range for the next control cycle: The system regenerates the executable control range for the next control cycle based on the execution feedback results, the current status of the equipment, and the field control rules. The process recording unit records the input status, executable control range, strategy preference information, equipment-level control instructions, and feedback results for each control cycle. The system thus forms a closed-loop control process from status acquisition, range generation, preference selection, authorized execution to feedback update. Step S1 includes: S1-1. Collect the equipment operating status of the current control cycle: The system will... An adjustable energy device in the current control cycle The current output is recorded as , will the The lower limit of the power output of each adjustable energy device is denoted as: The upper limit of device power is denoted as The upward adjustment of the climbing limit is recorded as... The positive value of adjusting the downward climbing limit is recorded as . The duration of the control cycle is recorded as The lower limit of the permissible power under the condition, calculated based on the equipment's state of charge, voltage, current, and temperature protection conditions, is denoted as... The maximum permitted power for the state is denoted as The lower and upper limits of the state-permitted power are used to uniformly convert the state of charge, voltage, current and temperature protection conditions into executable boundaries in the form of power, so that subsequent calculations can determine whether the equipment has safe adjustment capabilities within the same power dimension. The ramp limit represents the maximum change capability of adjustable energy equipment to increase or decrease output power within a control cycle. The state of charge is used to represent the proportion of the current remaining electricity of the energy storage device to the available capacity. S1-2. When calculating the allowable adjustment range within the current control cycle, the system does not directly use the upper and lower limits of the equipment's rated power as the selection range for the artificial intelligence model. Instead, it simultaneously considers the equipment power boundary, ramp-up capability boundary, and field condition permissible boundary, and selects the adjustment range that is satisfied by all three types of boundaries. Based on this processing, the system calculates the minimum allowable adjustment of the i-th adjustable energy device within the current control cycle t using the following formula. The system calculates the first [item] based on the same equipment boundary conditions. An adjustable energy device in the current control cycle Maximum allowable adjustment within The significance of the above two formulas is that they unify the power boundary of the equipment itself, the ramp capability boundary, and the field condition conversion boundary into the power adjustment amount that can be increased and decreased within the current control cycle. This allows the subsequent artificial intelligence model to read only the adjustment range that has met the field safety conditions. Through the above maximum and minimum value calculations, the system compresses multiple safety conditions into a continuous allowable adjustment range. Only the adjustment amount that falls into this range is allowed to enter the subsequent strategy preference ranking process. S1-3, Generate executable control segment: When Not greater than When the i-th adjustable energy device has a safe adjustment range within the current control cycle, the system encapsulates the i-th adjustable energy device, the minimum allowable adjustment amount, the maximum allowable adjustment amount, the current control cycle, and the set of failure conditions into an executable control segment. ,in, Indicates the first An adjustable energy device in the current control cycle The control segments that can be authorized for use within. This indicates the failure judgment conditions that the executable control segment must continuously meet within the current control cycle, including at least the execution deviation exceeding the limit condition, the communication delay exceeding the limit condition, and the field alarm trigger condition. The significance of this formula is that it binds the safety adjustment range and the failure judgment conditions together, so that the artificial intelligence model can only select control segments with time constraints and failure constraints, and cannot rewrite the equipment adjustment amount alone. The conventional approach typically involves an AI model generating a complete control strategy first, followed by backend rule verification or interception. However, this step avoids having the AI ​​model generate device-level control commands initially. Instead, the industrial energy management system first generates an executable control range based on the field equipment status, safety boundaries, operational limitations, and the current control cycle. This works by converting the safety requirements scattered across equipment limits, operational states, communication states, and protection conditions into a selectable control space. The AI ​​model can then only interact with control content that has already undergone field safety constraint processing. This step acts as a control entry point limiter in the entire solution, preventing the AI ​​model from accessing control objects exceeding the equipment's safety boundaries from the outset. Compared to post-event interception-based safety verification, this step moves safety constraints forward, before strategy generation, reducing the risk of out-of-bounds control from the source.

[0017] Step S2 includes: S2-1. Converting User Operating Objectives into Planned Deviation Correction Objectives: To avoid the AI ​​model directly generating device-level control commands based on natural language objectives, the system first converts the user operating objectives into planned deviation correction objectives that the industrial energy management system can quantify. These objectives represent the degree of deviation between the current total output and the planned output, and serve as a unified basis for the subsequent sequencing of executable control segments. The system will then use the current control cycle... The planned output is recorded as The total output measured on site will be recorded as The planned deviation for the current control cycle is calculated using the following formula. ,in, This formula represents the output deviation that needs to be corrected within the current control cycle. The significance of this formula is to convert the user-proposed economic operation, plan tracking, and scheduling correction targets into power deviation correction amounts that the field control system can recognize. Through this calculation, the abstract operation target input by the user is converted into a specific power deviation amount, so that the artificial intelligence model can only sort segments around the deviation correction effect instead of directly generating control commands for the field equipment. S2-2. Calculate the correction contribution of each executable control segment to the plan deviation: The system generates candidate adjustment amounts within the allowable adjustment interval corresponding to each executable control segment Cᵢ(t). The candidate adjustment values, along with their corresponding executable control segments, are then provided to the AI ​​model for ranking; the AI ​​model does not change... and Furthermore, no new equipment adjustments are generated outside the permitted adjustment range. The system calculates the contribution of candidate adjustments to the plan deviation using the following formula. ,in, This indicates the reduction in planned deviation after adopting candidate adjustment amounts. The significance of this formula is that it allows the ranking criteria of the artificial intelligence model to directly correspond to the planned deviation correction target in the industrial field, rather than outputting equipment commands that deviate from the field control boundary. S2-3, Generate policy preference information: The system will calculate the correction contributions of each executable control segment. and the corresponding executable control segment Input the AI ​​model and make it output a sequence of policy preferences. ,in, This indicates the ranking of executable control segments recommended by the artificial intelligence model within the current control cycle. to This represents the sorted adjustable energy device numbers. The significance of this formula lies in restricting the output of the artificial intelligence model to the sorting result, allowing the AI ​​model to express policy preferences rather than directly generating device-level control commands. The policy preference sequence... It only indicates the order of use of executable control segments recommended by the artificial intelligence model, and does not include the target output of the equipment, the switching actions of the equipment, or the execution format that can be directly sent to the field equipment; The innovation of this step lies in limiting the output of the artificial intelligence (AI) model. Instead of directly generating target output, switching actions, or execution commands, the AI ​​model only sorts and describes the already formed executable control segments. Its working principle is as follows: the system first converts the user's operational goals into quantifiable deviation correction goals that can be processed on-site. Then, it provides the AI ​​model with the effects of different executable control segments on these goals, allowing the AI ​​model to determine which control segments are more suitable for priority use. This step allows AI to participate in optimization without gaining direct control, enabling the AI ​​model's advantages to be reflected in sorting, filtering, and strategy selection, rather than directly controlling on-site equipment. Compared to conventional AI scheduling schemes where the model directly provides control quantities, this step limits the AI ​​output to a non-executable format, balancing strategy flexibility and industrial control security.

[0018] Step S3 includes: S3-1. Selecting the authorized control segment based on the policy preference sequence: The authorization execution and feedback correction module receives the policy preference sequence. Subsequently, instead of directly executing the policy preference sequence, the system reads executable control segments one by one according to the order of the policy preference sequence. For each executable control segment read, the system first confirms that the segment is still within the current control cycle and its failure condition set. Not triggered; only executable control segments that meet the above conditions are marked as authorized control segments. ,in, Indicates the first The formula defines a control segment that is safely authorized for use by an adjustable energy device within the current control cycle. The significance of this formula is that it clarifies that the authorized control segment can only come from the executable control range that has been generated in step S1, and the artificial intelligence model cannot add control content outside of this range. S3-2, Generating device-level control commands from authorized control segments: In authorized control segments Once determined, the system selects the actual adjustment amount from the candidate adjustment amounts corresponding to the authorized control segment. and ensure Still falling into to Within a defined range, based on the actual adjustment amount, the system generates a target output that the field equipment can recognize. ,in, Indicates from the authorized control segment The actual adjustment amount selected in the middle, Indicates that the document has been issued to the [number]th [number]. The target output of an adjustable energy device is defined by this formula, which transforms the strategy preferences recommended by the artificial intelligence model into a target output that the industrial field device can execute, while ensuring that the target output comes from the already authorized adjustment amount. It is not a control result directly issued by the artificial intelligence model, but a device execution target calculated by the security authorization side based on the authorized control segment; S3-3. Generate and issue device-level control commands: The system will... Equipment number and target output of each adjustable energy device Current control cycle and authorized control segment Together, generate device-level control commands ,in, This indicates the device-level control command sent to the field edge control device. The significance of this formula is to bind the authorized source and the target output, so that the field execution device can trace which authorized control segment each command comes from. This step is a crucial authorization process that transforms the AI ​​ranking results into field control actions. The ranking results provided by the AI ​​model themselves do not have direct execution capability. The system needs to reconfirm that the selected executable control segments are still within the current control cycle and that the corresponding failure conditions have not been triggered. Only then does the authorization execution and feedback correction module generate device-level control commands that the field equipment can recognize. Its working principle is that the device-level control commands are not directly issued by the AI ​​model, but rather recalculated and synthesized by the system based on the already authorized control segments, thus ensuring that each command can be traced back to the previously generated executable control range. This step plays a role in isolating control authority in the entire solution, ensuring that the AI ​​model can only affect the selection order of control segments and cannot bypass the industrial control system to directly change the state of the field equipment. Compared to conventional AI control architectures, this step clearly separates strategy recommendations and equipment execution, enabling better adaptation to existing energy management systems and field control interfaces.

[0019] Step S4 includes: S4-1, Acquisition of device execution feedback: In device-level control commands After being issued, the execution feedback unit collects the first... The measured output of an adjustable energy device at the next feedback moment The execution deviation is calculated according to the following formula. ,in, Indicates the first The measured output of an adjustable energy device at the next feedback moment. This formula represents the deviation between the actual output of the equipment and the target output. The significance of this formula lies in determining whether the equipment has completed the execution according to the authorized target output. S4-2. Determine if the executable control segment has failed: The system uses the execution deviation exceeding the limit condition and the communication delay exceeding the limit condition as the set of failure conditions. The specific judgment content is used to judge the authorized control segment. Whether there is still a safe basis for continuing execution, the system will record the execution deviation threshold as follows: The feedback communication delay is denoted as The allowed communication delay threshold is denoted as ,when Greater than At that time, the system determines the authorized control segment. Failure; when Greater than At that time, the system determines the authorized control segment. If an authorized control segment is deemed invalid, it will no longer be used as a source of authorization for subsequent control cycles. S4-3, Execution Failure Rollback Control: When at least one of the following conditions is met: execution deviation exceeds the allowable execution deviation threshold, or feedback communication delay exceeds the allowable communication delay threshold, the system determines the authorized control segment. The system has failed and will switch the actual control output to the preset safe output at the next feedback moment; if neither of the above two conditions is met, the system will maintain the original target output, and the system will... The preset safe output of each adjustable energy device is denoted as The actual control output at the next feedback moment is generated according to the following formula. ,in, Indicates the next feedback moment for the first The actual control output adopted by an adjustable energy device is defined by the following formula: when the execution deviation exceeds the allowable execution deviation threshold, the system switches to the safe output preset by the industrial control system; when the communication delay exceeds the allowable communication delay threshold, the system also switches to the safe output preset by the industrial control system; when both the execution deviation and the communication delay are within the allowable range, the system continues to use the authorized target output. This step addresses the issue of control commands losing their safety basis during execution due to equipment response, communication delays, or changes in field conditions. Even if a control segment meets safety conditions at the time of generation, the execution result of field equipment may still deviate from expectations due to load fluctuations, communication anomalies, insufficient equipment response, or alarm triggering. Therefore, the system needs to continue collecting execution feedback after the command is issued and determine whether the authorized control segment is still valid based on the feedback results. Its working principle is that the system does not consider authorized control commands as permanently valid, but rather uses execution deviations and communication status as continuous judgment criteria. Once an over-limit situation occurs, subsequent authorizations are revoked and the system switches to a preset safe control state. This step plays a dynamic protection role in the solution. Compared to the conventional method of only performing a verification before the command is issued, it can prevent expired or unsuitable AI recommendation results from continuing to affect equipment operation.

[0020] Step S5 includes: S5-1. Update the input state for the next control cycle based on the execution feedback: The system will update the measured output force obtained in step S4. Execution deviation and communication delay Forming a single device feedback status The system then summarizes the individual device feedback states of all adjustable energy devices into the state input set for the next control cycle. ,in This indicates the number of adjustable energy devices currently connected to the system. The significance of this formula is that the executable control range of the next control cycle directly inherits the execution result of the previous control cycle, thus avoiding the artificial intelligence model from continuing to participate in strategy selection based on the expired state. S5-2, Regenerate the executable control range for the next control cycle: The system will set the state inputs. Input the boundary conversion unit from step S1 and recalculate the minimum allowable adjustment for the next control cycle. and maximum allowable adjustment This generates the executable control segment for the next control cycle. This step reauthorizes each control cycle, and the executable control segments from the previous control cycle are not continued by default. S5-3. Record the closed-loop control process: The system summarizes all executable control segments generated within the current control cycle into a control segment set for the current control cycle t. k represents the number of executable control segments generated within the current control cycle, and all device-level control instructions actually issued within the current control cycle are aggregated into the control instruction set for the current control cycle t. This ensures that the process recording object covers the entire control chain from the generation of the executable scope to the issuance of device commands. 'm' represents the number of device-level control commands actually issued within the current control cycle, forming a set of control segments. and control command set Then, the system writes the process record value of the previous control cycle, the state input set of the current control cycle, the control segment set, the strategy preference sequence, the control instruction set, and the state input set of the next control cycle into the hash calculation process to obtain the process record value of the current control cycle t. ,in This represents the process record value from the previous control cycle. The hash calculation function is represented by the above recording method. The system can review afterward whether each device-level control instruction comes from the set of executable control segments generated in the current control cycle, and can confirm whether the artificial intelligence model always outputs the strategy preference sequence only within the set of executable control segments, thereby avoiding the artificial intelligence model from bypassing the executable control range to directly generate device control instructions.

[0021] This step transforms the execution results of the previous control cycle into the control basis for the next control cycle, enabling the invention to form a continuous closed loop rather than a one-time strategy generation process. After each control cycle, the system uses the actual equipment status, execution deviation, communication delay, and alarm status as input to update the executable control range for the next control cycle. Executable control segments generated in the previous cycle do not automatically continue. Its working principle lies in the fact that the field control environment is constantly changing. Only by incorporating actual execution feedback into the next round of control range generation can the AI ​​model ensure that it faces currently effective control content each time it participates in the sorting process. Simultaneously, the process recording unit continuously records the status input, executable control range, sorting results, equipment-level control commands, and feedback results, allowing for post-event verification of whether each equipment-level control command originates from an authorized control segment. This step plays a role in closed-loop updating and process traceability in the solution, enabling the system not only to generate commands safely but also to prove that the AI ​​model has not bypassed the executable control range to directly control the equipment.

[0022] Through the above steps, this invention establishes a control chain that first determines the safe and executable range, then performs artificial intelligence sorting, then generates device-level control instructions by the authorization module, and finally updates the control range based on execution feedback. The core of this chain lies not in simply improving the predictive ability of the artificial intelligence model itself, but in changing the way the artificial intelligence model is integrated into the industrial energy control system. This transforms the artificial intelligence model from a direct controller into a restricted strategy selector, ensuring that the industrial energy management system always retains the final command generation and execution control rights. Therefore, even if the output of the artificial intelligence model is uncertain, the system can still limit its influence boundaries through executable control range, authorized control segments, failure rollback, and closed-loop update mechanisms, thereby improving the safety, interpretability, and engineering feasibility of the industrial artificial intelligence energy control system.

[0023] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0024] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An energy management system for industrial AI control, characterized in that: include: The status acquisition and executable control range generation module is used to access the real-time operating status of the industrial energy management system, monitoring and data acquisition system, energy storage management equipment, and edge control equipment, and generate an executable control range that allows the artificial intelligence model to participate in the selection within the current control cycle based on the real-time operating status; the strategy preference generation module is used to receive the user's operating goal and the executable control range, and enable the artificial intelligence model to generate strategy preference information within the executable control range, wherein the strategy preference information is used to indicate the selection order and recommendation reasons of different executable control segments; The authorization execution and feedback correction module is used to generate device-level control instructions from the executable control range according to the policy preference information, receive feedback results after the device executes the instructions, and update the executable control range for the next control cycle based on the feedback results.

2. The energy management system for industrial AI control according to claim 1, characterized in that: The status acquisition and executable control range generation module includes a status access unit, a boundary conversion unit, and a control segment generation unit. The status access unit is used to acquire the current output, equipment operating limits, equipment charge status, communication status, and alarm status of the adjustable energy equipment. The boundary conversion unit is used to convert the equipment operating limits and on-site operating status into the power range that can be adjusted within the current control cycle. The control segment generation unit is used to combine the equipment object, the allowable adjustment range, the effective period, and the failure condition into an executable control segment. The strategy preference generation module includes a requirement parsing unit, a fragment sorting unit, and an output type restriction unit. The requirement parsing unit is used to convert the user-inputted running objective into an energy control objective. The fragment sorting unit is used to enable the artificial intelligence model to sort the executable control fragments according to the energy control objective. The output type restriction unit is used to restrict the artificial intelligence model to output only strategy preference information and not device-level control instructions. The authorized execution and feedback correction module includes an instruction synthesis unit, an edge distribution unit, an execution feedback unit, a failure rollback unit, and a process recording unit. The instruction synthesis unit is used to convert the authorized executable control segments into device-level control instructions. The edge distribution unit is used to distribute the device-level control instructions to field devices. The execution feedback unit is used to collect field feedback results after execution. The failure rollback unit is used to switch to a preset safety control state when the executable control segment fails. The process recording unit is used to record the executable control range, strategy preference information, device-level control instructions, and feedback results.

3. The energy management system for industrial AI control according to claim 2, characterized in that: The system performs the following steps: S1. Generate executable control range: The status acquisition and executable control range generation module acquires the real-time operating status of adjustable energy equipment on site and reads the equipment operating limits and safety control rules configured in the industrial energy management system. The system determines the direction and range that each device can be safely adjusted within the current control cycle based on the current equipment status. The system forms executable control segments by including the equipment objects that can be safely adjusted, the allowed adjustment content, the effective period, and the failure conditions, and summarizes them into executable control ranges. S2. Generate strategy preference information: The strategy preference generation module receives the user's running objective and the executable control range, and converts the user's running objective into a strategy selection task that the artificial intelligence model can understand. The artificial intelligence model must not directly output device-level control instructions, but can only sort and describe the executable control segments within the executable control range. The system thus obtains strategy preference information. S3. Generate and issue device-level control commands: After receiving the policy preference information, the authorization execution and feedback correction module reads the executable control segments in sequence according to the sorting results. The system only authorizes the executable control segments that are still in effect and have not triggered the failure condition, and converts the authorized executable control segments into device-level control commands that the field devices can recognize. The device-level control commands are sent to the corresponding field devices by the edge issuing unit. The artificial intelligence model does not participate in the direct issuance of the final command. S4. Collect execution feedback and handle failure situations: The execution feedback unit collects the output feedback, status feedback, communication feedback and alarm feedback after the field equipment is executed. The system will make a consistency judgment between the feedback results and the issued equipment-level control commands to confirm whether the equipment execution results are still in a safe control state. If the executable control segment fails, the failure rollback unit will revoke the subsequent authorization and switch the corresponding equipment to the preset safe control state. S5. Update the executable control range for the next control cycle: The system regenerates the executable control range for the next control cycle based on the execution feedback results, the current status of the equipment, and the field control rules. The process recording unit records the input status, executable control range, strategy preference information, equipment-level control instructions, and feedback results for each control cycle. The system thus forms a closed-loop control process from status acquisition, range generation, preference selection, authorized execution to feedback update.

4. An energy management system for industrial AI control as described in claim 3, Its features are: Step S1 includes: S1-1. Collect the equipment operating status of the current control cycle: The system will... An adjustable energy device in the current control cycle The current output is recorded as , the first The lower limit of the power output of each adjustable energy device is denoted as: The upper limit of device power is denoted as The upward adjustment of the climbing limit is recorded as... The positive value of adjusting the downward climbing limit is recorded as . The duration of the control cycle is recorded as The lower limit of the permissible power under the condition, calculated based on the equipment's state of charge, voltage, current, and temperature protection conditions, is denoted as... The maximum permitted power for the state is denoted as ; S1-2. Calculate the minimum allowable adjustment amount of the i-th adjustable energy device within the current control cycle t using the following formula. The system calculates the first [item] based on the same device boundary conditions. An adjustable energy device in the current control cycle Maximum allowable adjustment within ; S1-3, Generate executable control segment: When Not greater than When the system indicates that the i-th adjustable energy device has a safe adjustment range within the current control cycle, the system encapsulates the i-th adjustable energy device, the minimum allowable adjustment amount, the maximum allowable adjustment amount, the current control cycle, and the set of failure conditions into an executable control segment. ,in, Indicates the first An adjustable energy device in the current control cycle The control segments that can be authorized for use within. The failure judgment conditions that the executable control segment needs to continuously meet within the current control cycle include at least the execution deviation exceeding the limit condition, the communication delay exceeding the limit condition, and the on-site alarm trigger condition.

5. An energy management system for industrial AI control according to claim 4, characterized in that: Step S2 includes: S2-1. Convert user operational goals into planned deviation correction goals: Convert the current control cycle... The planned output is recorded as The total output measured on site will be recorded as The planned deviation for the current control cycle is calculated using the following formula. ,in, This indicates the output deviation that needs to be corrected within the current control cycle; S2-2, Calculate the correction contribution of each executable control segment to the plan deviation: The system in each executable control segment Candidate adjustment amounts are generated within the corresponding allowable adjustment range. The candidate adjustment values, along with their corresponding executable control segments, are then provided to the artificial intelligence model for sorting; the artificial intelligence model does not change... and Nor will it generate new equipment adjustments outside the allowed adjustment range. The system calculates the correction contribution of the candidate adjustments to the plan deviation according to the following formula. ,in, This indicates the reduction in planning deviation that can be achieved by adopting the aforementioned candidate adjustment amount; S2-3, Generate policy preference information: The system will calculate the correction contributions of each executable control segment. and the corresponding executable control segment Input the AI ​​model and make it output a sequence of policy preferences. ,in, This indicates the ranking of executable control segments recommended by the artificial intelligence model within the current control cycle. to This indicates the number of the adjustable energy device that has been sorted.

6. An energy management system for industrial AI control according to claim 5, characterized in that: Step S3 includes: S3-1. Selecting the authorized control segment based on the policy preference sequence: The authorization execution and feedback correction module receives the policy preference sequence. Subsequently, instead of directly executing the policy preference sequence, the system reads executable control segments one by one according to the order of the policy preference sequence. For each executable control segment read, the system first confirms that the segment is still within the current control cycle and its failure condition set. Not triggered; only executable control segments that meet the above conditions are marked as authorized control segments. ,in, Indicates the first A control segment that is safely authorized for use by an adjustable energy device during the current control cycle; S3-2, Generating device-level control commands from authorized control segments: In authorized control segments Once determined, the system selects the actual adjustment amount from the candidate adjustment amounts corresponding to the authorized control segment. and ensure Still falling into to Within the defined range, based on the actual adjustment amount, the system generates a target output that the field equipment can recognize. ,in, Indicates from the authorized control segment The actual adjustment amount selected in the middle, Indicates that it is sent to the first The target output of an adjustable energy device; S3-3, Generate and issue device-level control commands: The system will... Equipment number and target output of each adjustable energy device Current control cycle and authorized control segment Together, generate device-level control commands ,in, This indicates the device-level control command sent to the field edge control device.

7. An energy management system for industrial AI control according to claim 6, characterized in that: Step S4 includes: S4-1, Acquisition of device execution feedback: In device-level control commands After being issued, the execution feedback unit collects the data of the first... The measured output of an adjustable energy device at the next feedback moment The execution deviation is calculated according to the following formula. ,in, Indicates the first The measured output of an adjustable energy device at the next feedback moment. This indicates the deviation between the actual output of the equipment and the target output. S4-2. Determine if the executable control segment has failed: The system uses the execution deviation exceeding the limit condition and the communication delay exceeding the limit condition as the set of failure conditions. The specific judgment content is used to judge the authorized control segment. Whether there is still a safe basis for continuing execution, the system will record the execution deviation threshold as follows: The feedback communication delay is denoted as The allowed communication delay threshold is denoted as ,when Greater than At that time, the system determines the authorized control segment. Failure; when Greater than At that time, the system determines the authorized control segment. If an authorized control segment is deemed invalid, it will no longer be used as a source of authorization for subsequent control cycles. S4-3, Execution Failure Rollback Control: When at least one of the following conditions is met: execution deviation exceeds the allowable execution deviation threshold, or feedback communication delay exceeds the allowable communication delay threshold, the system determines the authorized control segment. It has failed, and the actual control output at the next feedback moment will be switched to the preset safe output; when neither the execution deviation exceeding the allowable execution deviation threshold nor the feedback communication delay exceeding the allowable communication delay threshold is met, the system maintains the original target output, and the system will... The preset safe output of each adjustable energy device is denoted as The actual control output at the next feedback moment is generated according to the following formula. ,in, Indicates the next feedback time for the first... The actual controlled output of each adjustable energy device.

8. An energy management system for industrial AI control according to claim 7, characterized in that: Step S5 includes: S5-1. Update the input state for the next control cycle based on the execution feedback: The system will update the measured output force obtained in step S4. Execution deviation and communication delay Forming a single device feedback status The system then summarizes the individual device feedback states of all adjustable energy devices into the state input set for the next control cycle. ,in This indicates the number of adjustable energy devices currently connected to the system; S5-2, Regenerate the executable control range for the next control cycle: The system will set the state inputs. Input the boundary conversion unit from step S1 and recalculate the minimum allowable adjustment for the next control cycle. and maximum allowable adjustment This generates the executable control segment for the next control cycle. ; S5-3. Record the closed-loop control process: The system summarizes all executable control segments generated within the current control cycle into a control segment set for the current control cycle t. k represents the number of executable control segments generated within the current control cycle, and all device-level control instructions actually issued within the current control cycle are aggregated into the control instruction set for the current control cycle t. This ensures that the process recording object covers the entire control chain from the generation of the executable scope to the issuance of device commands, where m represents the number of device-level control commands actually issued within the current control cycle, forming the control segment set. and the set of control commands Then, the system writes the process record value of the previous control cycle, the state input set of the current control cycle, the control segment set, the strategy preference sequence, the control instruction set, and the state input set of the next control cycle into the hash calculation process to obtain the process record value of the current control cycle t. ,in This represents the process record value from the previous control cycle. This represents the hash calculation function.

Images