Control instruction generation method and device for power distribution network and computer equipment

Abstract

The invention discloses a control instruction generation method and device for a power distribution network and computer equipment. The method comprises the steps that multiple information data sets of a power distribution network system are acquired, and the multiple information data sets comprise a power grid operation data set, an environment state data set and an image information data set; based on a multi-source sensing terminal, fusing the plurality of information data sets to obtain multi-source fusion data; based on an edge server, extracting information of the multi-source fusion data, and determining a security level and time delay of an extraction process; and determining a target control instruction of the power distribution network system based on the security level and the time delay. The technical problems of low control instruction generation efficiency and poor accuracy caused by function splitting of each link and low information interaction efficiency of a current power distribution network system are solved.

Description

Technical Field

[0001] The present invention relates to the field of electronic information technology, and in particular to a method, device and computer equipment for generating control instructions for a power distribution network. Background Art

[0002] With the transformation of energy structures and the rapid development of new power systems, the distribution network, as the core link connecting energy production and end users, faces increasing demands for intelligence, flexibility, and security. Traditional distribution networks often employ independently designed, layered deployment models for communication, perception, computing, control, and security protection systems. These systems are functionally fragmented and information exchange is inefficient, making them unable to cope with the challenges posed by the integration of high proportions of renewable energy, the dynamic response of diverse loads, and the large-scale application of new power electronic equipment. In particular, as new distribution networks evolve toward the collaborative interaction of "source-grid-load-storage," the current technical system faces multiple challenges, including insufficient communication and perception coordination capabilities, limited real-time computing and control, and a fragmented security protection system.

[0003] In recent years, breakthroughs in technologies such as multi-source sensing, secure communications, edge computing, and intelligent control have provided new insights for upgrading distribution networks. Building a technological framework that integrates ubiquitous sensing and communication, and coordinated computing and control, has become an inevitable trend in the development of smart distribution networks. However, existing research has largely focused on optimizing single technical areas, and has yet to develop an integrated architecture that deeply integrates sensing, communication, computing, control, and security. This has limited overall system performance and made it difficult to effectively generate control commands for distribution network systems.

[0004] To address the above-mentioned problems, no effective solutions have been proposed so far. Summary of the Invention

[0005] Embodiments of the present invention provide a method, apparatus, and computer device for generating control instructions for a distribution network, to at least address the current technical problems of low efficiency and poor accuracy in generating control instructions due to functional fragmentation and low efficiency of information exchange in various links of the distribution network system.

[0006] According to one aspect of an embodiment of the present invention, a method for generating control instructions for a distribution network is provided, comprising: acquiring multiple information data sets of a distribution network system, wherein the multiple information data sets include a power grid operation data set, an environmental status data set, and an image information data set; based on a multi-source perception terminal, fusing the multiple information data sets to obtain multi-source fused data; based on an edge server, extracting information from the multi-source fused data, and determining a security level and a delay of the extraction process; and determining a target control instruction of the distribution network system based on the security level and the delay.

[0007] Optionally, based on a multi-source perception terminal, multiple information data sets are fused to obtain multi-source fused data, including: standardizing the multiple information data sets to obtain multiple standard information data sets; determining the weights corresponding to the multiple standard information data sets based on the signal-to-noise ratio of the multiple information data sets; performing weighted summation on the multiple standard information data sets based on the weights to obtain initial fused data; encrypting the initial fused data to obtain encrypted fused data; transmitting the encrypted fused data to a target aggregation node, and fusing it with the encrypted fused data transmitted by other multi-source perception terminals to obtain multi-source fused data.

[0008] Optionally, based on the edge server, information of the multi-source fusion data is extracted, and the security level of the extraction process is determined, including: decrypting the multi-source fusion data to obtain the original fusion data; calculating the deviation between the multi-source fusion data and the original fusion data; and determining the security level based on the deviation.

[0009] Optionally, based on the edge server, information of multi-source fusion data is extracted and the delay of the extraction process is determined, including: decrypting the multi-source fusion data to obtain original fusion data; determining the processor operating cycle and processor operating frequency of the decryption process; determining the delay based on the original fusion data, processor operating cycle and processor operating frequency.

[0010] Optionally, after determining the target control instruction of the distribution network system based on the safety level and the time delay, the method includes: estimating the intrinsic entropy and differential entropy of the distribution network system after executing the target control instruction; determining the communication safety rate of the distribution network system based on the intrinsic entropy and the determinant of the covariance matrix; determining the cost lower limit corresponding to the target control instruction based on the differential entropy and the Riccati equation; determining the objective function based on the communication safety rate and the cost lower limit; and determining the risk assessment result of the target control instruction based on the objective function.

[0011] Optionally, based on the communication safety rate and the cost lower limit, the objective function is determined, including: determining the initial objective function based on the communication safety rate and the cost lower limit; determining the safety constraints of the distribution network system, wherein the safety constraints include delay constraints, deviation constraints and safety level constraints; based on the safety constraints, the initial objective function is optimized through the meta-agent to obtain the objective function.

[0012] According to another aspect of an embodiment of the present invention, a control instruction generation device for a distribution network is also provided, including: an acquisition module, used to acquire multiple information data sets of the distribution network system, wherein the multiple information data sets include a power grid operation data set, an environmental status data set and an image information data set; a fusion module, used to fuse the multiple information data sets based on a multi-source perception terminal to obtain multi-source fusion data; a first determination module, used to extract information from the multi-source fusion data based on an edge server, and determine the security level and delay of the extraction process; a second determination module, used to determine the target control instruction of the distribution network system based on the security level and delay.

[0013] According to another aspect of an embodiment of the present invention, a non-volatile storage medium is provided, which includes a stored program, wherein when the program is running, the device where the non-volatile storage medium is located is controlled to execute any one of the above-mentioned control instruction generation methods for the distribution network.

[0014] According to another aspect of an embodiment of the present invention, a computer device is provided. The computer device includes a processor, and the processor is used to run a program. When the program is run, any one of the above-mentioned control instruction generation methods for a power distribution network is executed.

[0015] According to yet another aspect of an embodiment of the present invention, a computer program product is provided, including a computer program. When the computer program is executed by a processor, any one of the above-mentioned control instruction generation methods for a power distribution network is implemented.

[0016] In an embodiment of the present invention, a control instruction generation method for a distribution network is adopted. By acquiring multiple information data sets of a distribution network system, wherein the multiple information data sets include a power grid operation data set, an environmental status data set, and an image information data set; based on a multi-source perception terminal, the multiple information data sets are fused to obtain multi-source fused data; based on an edge server, information of the multi-source fused data is extracted, and the security level and delay of the extraction process are determined; based on the security level and the delay, a target control instruction of the distribution network system is determined, thereby achieving the purpose of enabling the various modules of the distribution network system to work together to generate control instructions, thereby realizing the technical effect of improving the efficiency of the distribution network system in generating control instructions, and further solving the technical problems of low efficiency and poor accuracy in control instruction generation caused by functional fragmentation of various links in the current distribution network system and low efficiency of information interaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0018] Figure 1A hardware structure block diagram of a computer terminal for implementing a control instruction generation method for a power distribution network is shown;

[0019] Figure 2 is a flow chart of a method for generating control instructions for a power distribution network according to an embodiment of the present invention;

[0020] Figure 3 This is a flow chart of a novel method for operating a communication, perception, computing, and control security integrated architecture for a distribution network according to an optional embodiment of the present invention;

[0021] Figure 4 4 is a structural block diagram of a control instruction generating device for a power distribution network provided according to an embodiment of the present invention. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0023] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0024] According to an embodiment of the present invention, an embodiment of a method for generating control instructions for a power distribution network is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0025] The method embodiment provided in the first embodiment of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1The hardware structure block diagram of a computer terminal for implementing a control instruction generation method for a power distribution network is shown. Figure 1 As shown, the computer terminal 10 may include one or more (illustrated as 102a, 102b, ..., 102n in the figure) processors (the processor may include but is not limited to a microprocessor MCU or a programmable logic device FPGA and other processing devices), a memory 104 for storing data. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the BUS bus), a network interface, a power supply and / or a camera. It will be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 1 More or fewer components than shown, or with Figure 1 Different configurations shown.

[0026] It should be noted that the one or more processors and / or other data processing circuits described above may generally be referred to herein as "data processing circuitry." The data processing circuitry may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuitry may be a single, independent processing module, or may be incorporated in whole or in part into any of the other components of the computer terminal 10. As described in the embodiments of the present application, the data processing circuitry serves as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0027] The memory 104 can be used to store software programs and modules of application software, such as the program instruction / data storage device corresponding to the control instruction generation method for a power distribution network in an embodiment of the present invention. The processor executes the software programs and modules stored in the memory 104 to perform various functional applications and data processing, that is, to implement the control instruction generation method for a power distribution network of the above-mentioned application program. The memory 104 may include high-speed random access memory and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include memory remotely located relative to the processor, and these remote memories may be connected to the computer terminal 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0028] The display may be, for example, a touch screen liquid crystal display (LCD) that enables a user to interact with a user interface of the computer terminal 10 .

[0029] Figure 2FIG. 1 is a flow chart of a method for generating control instructions for a power distribution network according to an embodiment of the present invention. Figure 2 As shown, the method includes the following steps:

[0030] Step S201 : Acquire multiple information data sets of a power distribution network system, wherein the multiple information data sets include a power grid operation data set, an environmental status data set, and an image information data set.

[0031] In this step, the grid operation dataset is the core data set of the distribution network operation status, including but not limited to electrical parameters such as voltage, current, power, and frequency, as well as the status of electrical equipment, such as the position of circuit breakers and the status of relays. It may also include arc detection, power quality information (such as harmonics and voltage fluctuations), and fault location information. The environmental status dataset focuses on the external conditions of the distribution network area, such as meteorological parameters such as temperature, humidity, wind speed, and light intensity, as well as geographical environmental parameters such as soil moisture and vegetation coverage. The image information dataset mainly includes image data collected by visual sensors such as ground cameras and drone cameras. These images can be static photos or video streams. They are used to monitor the physical status of distribution network equipment, detect potential damage or failures (such as line breaks and equipment corrosion), and monitor environmental changes around the distribution network (such as construction activities and vegetation growth).

[0032] Step S202: Based on the multi-source perception terminal, multiple information data sets are fused to obtain multi-source fused data.

[0033] In this step, distribution network perception terminal management often faces problems such as rigid network topology (fixed aggregation nodes leading to resource waste), heterogeneous multimodal data (protocol differences hindering integration), and conflicts between security and timeliness. Current solutions often optimize network topology or data security in isolation, lacking a joint design of "topology-data-latency-security". Therefore, an efficient and secure management framework for multi-source perception terminals in the distribution network can be constructed. By optimizing the number of aggregation nodes, dynamic network topology planning can be achieved, and multiple information data sets of the distribution network system can be standardized and integrated. Key encryption is used to ensure data security, while also taking into account perception latency.

[0034] Step S203: extract information of multi-source fusion data based on the edge server, and determine the security level and delay of the extraction process.

[0035] In this step, traditional edge servers often separate data decryption, security assessment, and control decisions, resulting in accumulated computational delays (serial processing) or security vulnerabilities. Most current methods lack closed-loop modeling of the "decryption-computation-control" link, making it difficult to balance timeliness and security. Therefore, encrypted data features can be extracted using decryption keys, computational security can be assessed based on these features, and data processing complexity can be quantified to calculate task latency. Finally, control instructions can be generated by combining security level and latency, achieving a dynamic trade-off between security perception and real-time control, supporting efficient and reliable decision-making at the edge layer.

[0036] Furthermore, when transmitting data between multi-source sensing terminals and edge servers, traditional anti-eavesdropping schemes rely on fixed channel models or ignore the characteristics of short packets, resulting in distorted secure communication rate estimates (overestimating legitimate links or underestimating eavesdropping capabilities). Current research often relies on idealized channel assumptions (e.g., ignoring the dynamic changes in path loss), making it difficult to adapt to the complex electromagnetic environment of distribution networks. Therefore, to address the secure transmission requirements of distribution network communications, a path loss model can be established to derive the uplink and downlink channel gains. The legitimate communication rate and eavesdropper rate can then be calculated under the short packet transmission mechanism. Ultimately, the secure communication rate can be used to quantify anti-eavesdropping capabilities.

[0037] Specifically, assuming that during the execution of L tasks, the communication bandwidth is defined as B, and the communication path loss is defined as:

[0038]

[0039] Where PL is the path loss, P los is the line of sight probability, and is the fading margin reserved for shadow fading, clutter is the clutter effect, and the specific values ​​of these parameters depend on the actual conditions. η represents the free space path loss, which is the carrier frequency f l and communication distance d l Function:

[0040] The antenna gain formula can be calculated as:

[0041]

[0042] Among them, g1(r l ) and g3(r l ) are the first-order and third-order Bessel functions of the first kind, H max is the maximum antenna gain, r l It is given by:

[0043]

[0044] Among them, θl is the deviation angle between the two communicating parties (with the edge server as the origin), θ 3dB is the single-sided half-power beam width when the bandwidth is 3dB.

[0045] Then the uplink and downlink channel gains are calculated as:

[0046]

[0047] In order to ensure the timeliness of communication, short data communication can be selected. In this case, the communication rate of the system is expressed as:

[0048]

[0049] in, is the signal-to-noise ratio of legitimate communication, p l is the transmitted power, σ 2 The noise of legitimate communication, n l is the communication data packet block length, V(·) represents the channel dispersion, Q -1 (·) represents the inverse Q function, ε l Indicates the probability of communication block errors due to short packet transmission.

[0050] Since malicious eavesdroppers are unavoidable in actual distribution network operation and may eavesdrop on communication information, the communication rate of the eavesdropper can be calculated by the same logic as follows:

[0051]

[0052] Among them, h e is the gain of the eavesdropping channel, is the eavesdropping noise power. The secure communication rate of the system can be expressed as R s =R c -R e .

[0053] Step S204: determining a target control instruction of the distribution network system based on the security level and the time delay.

[0054] In this step, the target control instructions are generated based on the current security level and latency requirements. This generation process is not simply based on real-time data feedback, but rather incorporates considerations of the overall system security. The security level reflects the current security status of the system, including the security of data transmission, the security of the computational process, and the potential security risks posed by control actions. Latency refers to the time elapsed from data decryption to the execution of the control instructions. This ensures that control instructions can be issued and executed in a timely manner before the grid status changes, thereby adapting to the grid's rapidly changing needs.

[0055] Through the above steps, the purpose of making the various modules of the distribution network system work together to generate control instructions is achieved, thereby achieving the technical effect of improving the efficiency of the distribution network system in generating control instructions, and further solving the current technical problems of low efficiency and poor accuracy in control instruction generation caused by the functional fragmentation of various links in the distribution network system and low efficiency of information exchange.

[0056] As an optional embodiment, based on a multi-source perception terminal, multiple information data sets are fused to obtain multi-source fused data, including: standardizing the multiple information data sets to obtain multiple standard information data sets; determining the weights corresponding to the multiple standard information data sets based on the signal-to-noise ratio of the multiple information data sets; performing weighted summation on the multiple standard information data sets based on the weights to obtain initial fused data; encrypting the initial fused data to obtain encrypted fused data; transmitting the encrypted fused data to a target aggregation node, and fusing it with the encrypted fused data transmitted by other multi-source perception terminals to obtain multi-source fused data.

[0057] Optionally, the perception data set of the kth multi-source perception terminal can be defined as a power grid operation data set: Environmental status dataset and image information datasets I, J, and M are the dimensions of the three datasets. These are subsets of the collected data corresponding to the three datasets. In order to enhance the robustness of the perception data, the data of the three datasets can be standardized:

[0058]

[0059] in, is the standard information data set. Specifically, when y is g, is the standardized power grid operation data set; when y is e, is a standardized environmental state data set; when y is p, is a standardized image information dataset. The information data set perceived by the sensing terminal includes the power grid operation data set, the environmental status data set and the image information data set. is the mean standard data, To reflect the standard deviation of data fluctuations, ε is a very small positive number for numerical stability. Then, the weight ω can be assigned according to the importance and signal-to-noise ratio of each data set. y , where the weight vector satisfies The initial fusion data S k It is defined as the weighted sum:

[0060]

[0061] In order to ensure the security of the perception data, the sink node broadcasts its ID and data to its neighboring nodes and transmits it to the edge server, and assigns a key Ξ to each multi-source perception terminal. s , the data encryption calculation formula is: Each multi-source perception terminal aggregates its encrypted data to the adjacent target aggregation node, which further integrates the encrypted data. The formula is expressed as: K / k1 represents the number of multi-source sensing terminals managed by the aggregation node. To ensure the efficiency of sensing, the time for data encryption and fusion is defined as:

[0062]

[0063] in, and The CPU cycles required to encrypt one bit of data at the sink node and other multi-source sensing terminals, The CPU frequency allocated when encrypting data. The time required for each sensing terminal to aggregate its encrypted data to the adjacent target sink node.

[0064] In addition, for the number of sink nodes, in order to facilitate the management of K multi-source sensing terminals, k1 sink nodes can be selected according to the actual distribution network area to aggregate the sink nodes as the center and the radius is r. k Therefore, the optimal number of aggregation nodes is selected according to the actual distribution network operation situation. The formula is:

[0065]

[0066] in, Indicates the power parameter in the free space data transmission direction, It represents the actual power parameter under the influence of multiple factors, and D represents the range that needs to be perceived in actual distribution network operation.

[0067] As an optional embodiment, based on the edge server, information of multi-source fusion data is extracted and the security level of the extraction process is determined, including: decrypting the multi-source fusion data to obtain the original fusion data; calculating the deviation between the multi-source fusion data and the original fusion data; and determining the security level based on the deviation.

[0068] Optionally, after receiving the multi-source fusion data, the edge server performs information processing, i.e., decryption, to extract information from the multi-source perception data. The information extraction process can be described as follows:

[0069] S * =Ξ c ×S

[0070] Among them, c is a decryption key, c ×S represents the task-related information extracted from the sensory data, and also represents the information contained in the dataset, and then generates control actions based on the information By calculating the deviation S of the information extraction process * -S, to measure the safety level of the calculation.

[0071] As an optional embodiment, based on the edge server, information of multi-source fusion data is extracted and the delay of the extraction process is determined, including: decrypting the multi-source fusion data to obtain the original fusion data; determining the processor operating cycle and processor operating frequency of the decryption process; and determining the delay based on the original fusion data, the processor operating cycle and the processor operating frequency.

[0072] Optionally, decrypting the multi-source fusion data is performed in the same manner as in determining the security level, and the calculation delay can be achieved using the following formula:

[0073]

[0074] Among them, α l Indicates the difficulty of data processing (the CPU cycle required to process one bit of data), that is, the processor operating cycle, f l cpu It indicates the CPU frequency allocated when the lth task is executed, that is, the processor operating frequency, and S is the original fusion data obtained by decryption.

[0075] As an optional embodiment, after determining the target control instruction of the distribution network system based on the security level and delay, it includes: estimating the intrinsic entropy and differential entropy of the distribution network system after executing the target control instruction; determining the communication safety rate of the distribution network system based on the intrinsic entropy and the covariance matrix determinant; determining the cost lower limit corresponding to the target control instruction based on the differential entropy and the Riccati equation; determining the objective function based on the communication safety rate and the cost lower limit; and determining the risk assessment result of the target control instruction based on the objective function.

[0076] Alternatively, traditional assessment models for control command security risks often rely on single-dimensional indicators and ignore the multimodal coupling effect (the propagation of perception errors to control risks). Therefore, a comprehensive assessment framework can be proposed. Based on intrinsic entropy and the determinant of the covariance matrix, a stability condition for the communication security rate is set. The cost lower bound of the control command is solved through differential entropy and the Riccati equation. Then, an objective function is constructed. The standard data set and weight coefficients are combined to quantify the risk impact of perception-communication-computation on control security, achieving interpretable modeling and dynamic early warning of multi-dimensional risks.

[0077] Specifically, to ensure stable operation of the system, the safe communication rate must meet the following conditions:

[0078] R s ≥log2|det(Φ l )|

[0079] Among them, log2|det(Φ l )| is the intrinsic entropy that reveals the level of instability of the system, det(Φ l ) is the calculation matrix Φ l Under the condition of system stability, the communication security rate also constructs the lower limit of the control behavior value cost function:

[0080]

[0081] in, g(·) is the differential entropy function, Ψ l is the covariance matrix of the process noise during task execution, W l and E l This can be obtained through the following Riccati equation:

[0082] W l =A l +Φ l T (W l -E l )Φ l

[0083] E l =W l T Γ l (B l +Γ l W l Γ l ) -1 Γ l T W l

[0084] in is the weighted matrix of the value between the equilibrium state deviation and the control input, It is the influence matrix of quantitative control behavior on state evolution. According to the above formula, the objective function of optimization, namely the rate value cost function, can be defined to quantify the relationship between the extreme control value cost and the safe communication rate:

[0085]

[0086] In order to evaluate the security of control, a security risk assessment function of control can be constructed based on the objective function, where the security of control is affected by perception, communication and calculation. The formula is expressed as:

[0087]

[0088] Where, A standard data set for error-free encryption of sensing terminals, and are weight coefficients respectively, satisfying Based on the above security risk assessment function, the final risk assessment result can be obtained.

[0089] As an optional embodiment, the objective function is determined based on the communication safety rate and the cost lower limit, including: determining the initial objective function based on the communication safety rate and the cost lower limit; determining the safety constraints of the distribution network system, wherein the safety constraints include delay constraints, deviation constraints and safety level constraints; based on the safety constraints, the initial objective function is optimized through the meta-agent to obtain the objective function.

[0090] Alternatively, because traditional reinforcement learning algorithms face challenges with high-dimensional constraints and poor adaptability to dynamic environments in distribution networks, a single agent struggles to coordinate multi-objective conflicts. Current meta-learning approaches often focus on generalizing a single task and fail to address the hierarchical control requirements of distribution networks. Therefore, a multi-agent meta-reinforcement learning framework can be designed. Local agents (LAs) are deployed to perform real-time local adjustments, while meta-agents (MAs) optimize global policies through meta-learning, dynamically optimizing the system's objective function to achieve safety constraints. Through hierarchical interaction, this framework achieves joint optimization of communication security, perception security, computational latency, computational security, and control risk, supporting adaptive collaborative decision-making within the system under complex constraints.

[0091] Specifically, the multi-agent reinforcement learning algorithm is used to optimize the objective function while meeting security constraints. For example, the encryption and fusion time of the perception data, the calculation time, and the security risk of the control are all less than the set threshold, which can be expressed as R <R thre ,(S * -S) thre ,in and are the maximum delay thresholds for perception and calculation, S thre is the data deviation threshold during calculation, R thre The multi-agent meta-reinforcement learning algorithm framework is divided into two types of agents: one is the local agent (LA) located at the perception terminal and actuator, which is responsible for adjusting the local strategy; the other is the meta-agent (MA) located at the edge server, which is responsible for optimizing the global scheduling strategy.​

[0092] The state spaces of LA and MA can be defined as: is the state space of the multi-source sensing terminal, where It is a standard information dataset for power grid operation, a standard information dataset for environmental status, and a standard information dataset for image information. is the encrypted data of the sensing terminal, t sens The time for encrypting and fusing the sensing terminal data. is the state space of the actuator, where u l,t Control actions produced for edge servers, LOSS l is the cost function of the control behavior value. is the state space of the edge server, where Calculate the latency for the edge server, R s is the safe rate of communication, x l,t+1 is the system state, R is the control safety risk assessment function. The corresponding action spaces are k1 is the number of sink nodes selected by the multi-source sensing terminal according to the actual distribution network area, ω y is the allocation weight of perception data fusion, s The key assigned to each multi-source sensing terminal, The CPU frequency allocated when encrypting data. and The CPU cycles required to encrypt one bit of data for the sink node and other multi-source sensing terminals respectively; in is the weighted matrix of the value between the equilibrium state deviation and the control input, A matrix that quantifies the effect of control actions on state evolution; h in l is the communication channel gain of uplink and downlink, h e is the gain of the eavesdropping channel, p l is the power transmitted during communication, f l cpu Indicates the CPU frequency allocated to the edge server when executing the lth task, α l represents the data processing difficulty of the edge server (the CPU cycles required to process one bit of data), c is the edge server decryption key.

[0093] In order to measure the benefits of the agent's actions, the agent's reward function is defined as: γ1, γ2 and γ3 are weight coefficients. The state value function can be defined as and action-value function They are:

[0094]

[0095] Where υ is the discount factor, Indicates that the action is Status is The reward function when the policy is π, where π is the parameter θ χ The deep neural network of χ∈{sens,exec,comp} corresponds to three types of agents, The function is applicable to LA and MA. Just replace the corresponding actions and states with the actions and states of the agent. Finally, update the neural network parameters:

[0096]

[0097] Among them, l is the lth execution of the task, and a total of L times are executed. is the meta-learning step size.

[0098] As an optional embodiment, a new distribution network communication perception computing control security integrated architecture and a corresponding operation method are also provided. Figure 3 This is a flow chart of a novel distribution network communication perception computing control security integrated architecture operation method provided according to an optional embodiment of the present invention. Figure 3 As shown in Figure 1, this architecture consists of edge servers, multi-source sensing terminals, and actuators. Assuming linear time-invariance, the system's evolution equations during task execution are generated, and a cost-benefit function for control actions is defined to measure the value of the system's control strategy. Traditional distribution network architectures, due to the isolated design of communication, sensing, computing, control, and security modules, suffer from inefficient information exchange, delayed dynamic response, and insufficient multi-objective optimization capabilities, making them difficult to cope with complex scenarios such as renewable energy access and sudden load changes. While current solutions incorporate edge computing and multi-source sensing, their functional fragmentation, overly simplified models, and static cost function design lead to low resource utilization and poor control strategy adaptability. To address this, a new integrated communication-perception-computing-control-security architecture can be constructed. This architecture leverages edge servers to achieve decentralized computing power and real-time decision-making. Multi-source sensing terminals fuse panoramic data, linking actuators to form closed-loop control. Assuming linear time-invariance, the system's evolution equations during task execution are generated, and a cost-benefit function for control actions is defined to measure the value of the system's control strategy. This architecture deeply integrates physical properties with mathematical modeling, breaking through the limitations of traditional modularization.

[0099] Specifically, the architecture for a new integrated communication, perception, computing, and control technology for distribution networks is first constructed. This architecture comprises an edge server with communication and computing modules responsible for communication and information processing; K multi-source perception terminals responsible for sensing multi-source data from the distribution network, including grid operation data (such as voltage, current, and arcing), environmental data (such as temperature and humidity), and image information (such as from ground cameras and drone cameras); and M actuators to execute control tasks. During task execution, the multi-source perception terminals collect raw data and upload it to the edge server via an uplink. Based on the sensor data, the edge server generates control commands using inference computing and transmits them to the actuators via a downlink. The actuators then take actions, and the results of the actions are fed back to the edge server, forming a closed loop for the task.

[0100] Assume that the lth task is executed at time t. The evolution equation of the overall system generated by the edge server during task execution is:

[0101] x l,t+1 =Φ l x l,t +Γ l u l,t +v l,t

[0102] in, Indicates the system status, Indicates a control action. Represents process noise, whose covariance matrix is ​​expressed as Ψ l , represents a real matrix of dimension n×1, quantifies the intrinsic dynamics of the system in the absence of external intervention, The impact of control behavior on state evolution is quantified. In order to measure the value of control behavior, a value cost function LOSS can be defined l for:

[0103]

[0104] Among them, limsup(·) is the upper limit operation, represents mean calculation, is the weighted matrix of the value between the equilibrium state deviation and the control input, a T represents the transpose calculation, and L is the Lth task execution.

[0105] The above-mentioned new distribution network communication, perception, computing, control and security integrated architecture integrates communication, perception, computing, control and security mechanisms into a linear time-invariant system through the collaboration of edge servers, multi-source perception terminals and actuators, solving the problems of low efficiency and poor coordination caused by module dispersion in traditional technologies; through standardized multimodal data fusion, key encryption and secure communication rate quantification, combined with perception delay, computing delay and security indicators, a dynamic risk assessment model is constructed to improve the comprehensive defense capability against complex attacks and abnormal conditions; the secure communication rate design under the short data packet transmission mechanism quantifies the anti-eavesdropping capability, combined with key encryption and multi-source perception terminal security management framework, it reduces perception delay while ensuring data security; edge computing is used to process information and generate control strategies, reducing cloud dependence, communication overhead and computing delay, while ensuring the security of the computing process through secure computing indicators, and using multi-intelligent element reinforcement learning algorithms to optimize the overall system operation efficiency while ensuring system security. The multi-agent meta-reinforcement learning algorithm used is a relatively new and high-performance algorithm in the field of artificial intelligence. It not only overcomes the problem that traditional deep reinforcement learning relies on a large amount of scene data to train a single strategy and is difficult to adapt to the dynamically changing environment of the distribution network, but also the meta-agent extracts common strategies through cross-task learning, and the local agent quickly adapts to new scenarios, significantly improving the generalization ability and accelerating strategy convergence.

[0106] It should be noted that for the aforementioned method embodiments, for simplicity of description, they are all expressed as a series of action combinations. However, those skilled in the art should be aware that the present invention is not limited by the order of the actions described, because according to the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in this specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0107] Through the description of the above embodiments, those skilled in the art can clearly understand that the control instruction generation method for the power distribution network according to the above embodiment can be implemented by means of software plus the necessary general hardware platform. Of course, it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in each embodiment of the present invention.

[0108] According to an embodiment of the present invention, there is also provided a device for implementing the above-mentioned method for generating control instructions for a power distribution network. Figure 4FIG. 1 is a structural block diagram of a control instruction generating device for a power distribution network according to an embodiment of the present invention. Figure 4 As shown, the device includes: an acquisition module 41, a fusion module 42, a first determination module 43 and a second determination module 44. The device is described below.

[0109] The acquisition module 41 is configured to acquire multiple information data sets of the power distribution network system, wherein the multiple information data sets include a power grid operation data set, an environmental status data set, and an image information data set.

[0110] The fusion module 42 is connected to the acquisition module 41 and is used to fuse multiple information data sets based on the multi-source perception terminal to obtain multi-source fused data.

[0111] The first determination module 43 is connected to the fusion module 42 and is used to extract information of multi-source fusion data based on the edge server and determine the security level and delay of the extraction process.

[0112] The second determining module 44 is connected to the first determining module 43 and is configured to determine a target control instruction for the distribution network system based on the security level and the time delay.

[0113] It should be noted that the acquisition module 41, fusion module 42, first determination module 43, and second determination module 44 described above correspond to steps S201 to S204 in the embodiment. The examples and application scenarios implemented by the various modules and corresponding steps are the same, but are not limited to the contents disclosed in the above embodiment. It should be noted that the above modules, as part of the device, can be run in the computer terminal 10 provided in the embodiment.

[0114] An embodiment of the present invention may provide a computer device. Optionally, in this embodiment, the computer device may be located in at least one of a plurality of network devices in a computer network. The computer device includes a memory and a processor.

[0115] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the control instruction generation method and device for the power distribution network in the embodiments of the present invention. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, implementing the above-mentioned control instruction generation method for the power distribution network. The memory may include a high-speed random access memory and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory may further include a memory remotely located relative to the processor, and these remote memories may be connected to the computer terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0116] The processor can call the information and application programs stored in the memory through the transmission device to perform the following steps: obtain multiple information data sets of the distribution network system, wherein the multiple information data sets include a power grid operation data set, an environmental status data set, and an image information data set; based on the multi-source perception terminal, the multiple information data sets are integrated to obtain multi-source integrated data; based on the edge server, the information of the multi-source integrated data is extracted, and the security level and delay of the extraction process are determined; based on the security level and delay, the target control instructions of the distribution network system are determined.

[0117] Optionally, the processor may also execute the program code of the following steps: based on a multi-source perception terminal, fusing multiple information data sets to obtain multi-source fused data, including: standardizing multiple information data sets to obtain multiple standard information data sets; determining the weights corresponding to each of the multiple standard information data sets based on the signal-to-noise ratio of the multiple information data sets; performing weighted summation on the multiple standard information data sets based on the weights to obtain initial fused data; encrypting the initial fused data to obtain encrypted fused data; transmitting the encrypted fused data to a target aggregation node, and fusing it with the encrypted fused data transmitted by other multi-source perception terminals to obtain multi-source fused data.

[0118] Optionally, the above-mentioned processor can also execute the program code of the following steps: based on the edge server, extract the information of the multi-source fusion data, and determine the security level of the extraction process, including: decrypting the multi-source fusion data to obtain the original fusion data; calculating the deviation between the multi-source fusion data and the original fusion data; and determining the security level based on the deviation.

[0119] Optionally, the above-mentioned processor can also execute the program code of the following steps: based on the edge server, extract the information of the multi-source fusion data, and determine the delay of the extraction process, including: decrypting the multi-source fusion data to obtain the original fusion data; determining the processor operating cycle and processor operating frequency of the decryption process; determining the delay based on the original fusion data, the processor operating cycle and the processor operating frequency.

[0120] Optionally, the above-mentioned processor can also execute the program code of the following steps: after determining the target control instruction of the distribution network system based on the security level and delay, including: estimating the intrinsic entropy and differential entropy of the distribution network system after executing the target control instruction; determining the communication safety rate of the distribution network system based on the intrinsic entropy and the determinant of the covariance matrix; determining the cost lower limit corresponding to the target control instruction based on the differential entropy and the Riccati equation; determining the objective function based on the communication safety rate and the cost lower limit; and determining the risk assessment result of the target control instruction based on the objective function.

[0121] Optionally, the processor may also execute the program code of the following steps: determining the objective function based on the communication safety rate and the cost lower limit, including: determining the initial objective function based on the communication safety rate and the cost lower limit; determining the safety constraints of the distribution network system, wherein the safety constraints include delay constraints, deviation constraints and safety level constraints; based on the safety constraints, optimizing the initial objective function through the meta-agent to obtain the objective function.

[0122] An embodiment of the present invention provides a method for generating control instructions for a distribution network. The method obtains multiple information data sets from a distribution network system, including a power grid operation data set, an environmental status data set, and an image information data set; fuses the multiple information data sets based on a multi-source sensing terminal to obtain multi-source fused data; extracts information from the multi-source fused data based on an edge server, and determines the security level and latency of the extraction process; and determines target control instructions for the distribution network system based on the security level and latency. This method achieves the goal of enabling the various modules of the distribution network system to work together to generate control instructions, thereby achieving the technical effect of improving the efficiency of the distribution network system in generating control instructions. This further solves the current technical problem of low efficiency and poor accuracy in control instruction generation caused by the functional fragmentation of various links in the distribution network system and low information exchange efficiency.

[0123] A person skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing the hardware related to the terminal device through a program, and the program can be stored in a non-volatile storage medium, which may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0124] The embodiment of the present invention further provides a non-volatile storage medium. Optionally, in this embodiment, the non-volatile storage medium can be used to store program codes executed by the control instruction generation method for a power distribution network provided in the embodiment.

[0125] Optionally, in this embodiment, the non-volatile storage medium may be located in any computer terminal in a computer terminal group in a computer network, or in any mobile terminal in a mobile terminal group.

[0126] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for executing the following steps: obtaining multiple information data sets of the distribution network system, wherein the multiple information data sets include a power grid operation data set, an environmental status data set, and an image information data set; based on a multi-source perception terminal, fusing the multiple information data sets to obtain multi-source fused data; based on an edge server, extracting information from the multi-source fused data, and determining the security level and delay of the extraction process; and determining the target control instructions of the distribution network system based on the security level and delay.

[0127] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for executing the following steps: based on a multi-source perception terminal, fusing multiple information data sets to obtain multi-source fused data, including: standardizing multiple information data sets to obtain multiple standard information data sets; based on the signal-to-noise ratio of the multiple information data sets, determining the weights corresponding to each of the multiple standard information data sets; based on the weights, performing weighted summation on the multiple standard information data sets to obtain initial fused data; encrypting the initial fused data to obtain encrypted fused data; transmitting the encrypted fused data to a target aggregation node, and fusing it with the encrypted fused data transmitted by other multi-source perception terminals to obtain multi-source fused data.

[0128] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for performing the following steps: based on the edge server, extracting information of the multi-source fusion data, and determining the security level of the extraction process, including: decrypting the multi-source fusion data to obtain the original fusion data; calculating the deviation between the multi-source fusion data and the original fusion data; and determining the security level based on the deviation.

[0129] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for performing the following steps: based on the edge server, extracting information of the multi-source fusion data, and determining the delay of the extraction process, including: decrypting the multi-source fusion data to obtain the original fusion data; determining the processor operating cycle and processor operating frequency of the decryption process; determining the delay based on the original fusion data, the processor operating cycle and the processor operating frequency.

[0130] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for executing the following steps: after determining the target control instruction of the distribution network system based on the security level and the delay, including: estimating the intrinsic entropy and differential entropy of the distribution network system after executing the target control instruction; determining the communication safety rate of the distribution network system based on the intrinsic entropy and the determinant of the covariance matrix; determining the cost lower limit corresponding to the target control instruction based on the differential entropy and the Riccati equation; determining the objective function based on the communication safety rate and the cost lower limit; and determining the risk assessment result of the target control instruction based on the objective function.

[0131] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for performing the following steps: determining an objective function based on a communication safety rate and a cost lower limit, including: determining an initial objective function based on a communication safety rate and a cost lower limit; determining safety constraints of the distribution network system, wherein the safety constraints include delay constraints, deviation constraints, and safety level constraints; and optimizing the initial objective function through a meta-agent based on the safety constraints to obtain an objective function.

[0132] An embodiment of the present invention also provides a computer program product, including a computer program. Optionally, in this embodiment, when the computer program is executed by a processor, it can achieve: obtaining multiple information data sets of the distribution network system, wherein the multiple information data sets include a power grid operation data set, an environmental status data set, and an image information data set; based on a multi-source perception terminal, fusing the multiple information data sets to obtain multi-source fused data; based on an edge server, extracting information from the multi-source fused data, and determining the security level and delay of the extraction process; based on the security level and delay, determining the target control instructions of the distribution network system.

[0133] The serial numbers of the above embodiments of the present invention are for description only and do not represent the advantages or disadvantages of the embodiments.

[0134] In the above embodiments of the present invention, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0135] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only exemplary. For example, the division of the units can be a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0136] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0137] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0138] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a non-volatile storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk, and other media that can store program code.

[0139] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A control instruction generation method for a distribution network, characterized in that: include: Acquire multiple information data sets of a power distribution network system, wherein the multiple information data sets include a power grid operation data set, an environmental status data set, and an image information data set; Based on a multi-source sensing terminal, the multiple information data sets are fused to obtain multi-source fused data; Extracting information of the multi-source fusion data based on the edge server and determining the security level and latency of the extraction process; Based on the safety level and the time delay, a target control instruction of the power distribution network system is determined.

2. The method according to claim 1, characterized in that The method of fusing the plurality of information data sets based on the multi-source sensing terminal to obtain multi-source fused data includes: Standardizing the multiple information data sets to obtain multiple standard information data sets; Determining weights corresponding to the plurality of standard information data sets based on signal-to-noise ratios of the plurality of information data sets; Based on the weights, performing weighted summation on the multiple standard information data sets to obtain initial fusion data; Encrypting the initial fused data to obtain encrypted fused data; The encrypted fused data is transmitted to a target sink node and fused with the encrypted fused data transmitted by other multi-source perception terminals to obtain the multi-source fused data.

3. The method according to claim 1, characterized in that Extracting information of the multi-source fusion data based on the edge server and determining the security level of the extraction process include: Decrypting the multi-source fused data to obtain original fused data; Calculating the deviation between the multi-source fused data and the original fused data; Based on the deviation, the safety level is determined.

4. The method according to claim 1, wherein Extracting information of the multi-source fusion data based on the edge server and determining the delay of the extraction process includes: Decrypting the multi-source fused data to obtain original fused data; Determine the processor operating cycle and processor operating frequency of the decryption process; The time delay is determined based on the original fusion data, the processor operation cycle, and the processor operation frequency.

5. The method according to claim 1, characterized in that After determining the target control instruction of the distribution network system based on the security level and the time delay, the method further includes: estimating the intrinsic entropy and differential entropy of the power distribution network system after executing the target control instruction; Determining a communication security rate of the distribution network system based on the intrinsic entropy and the covariance matrix determinant; Determining a lower limit of the cost corresponding to the target control instruction based on the differential entropy and the Riccati equation; Determining an objective function based on the communication security rate and the cost lower bound; Based on the objective function, a risk assessment result of the target control instruction is determined.

6. The method according to claim 5, characterized in that The determining of the objective function based on the communication security rate and the cost lower limit includes: Determining an initial objective function based on the communication security rate and the cost lower bound; Determining safety constraints of the power distribution network system, wherein the safety constraints include delay constraints, deviation constraints, and safety level constraints; Based on the safety constraints, the initial objective function is optimized by a meta-agent to obtain the objective function.

7. A control instruction generating device for a power distribution network, characterized in that: include: An acquisition module, configured to acquire a plurality of information data sets of a power distribution network system, wherein the plurality of information data sets include a power grid operation data set, an environmental status data set, and an image information data set; A fusion module, configured to fuse the plurality of information data sets based on a multi-source sensing terminal to obtain multi-source fused data; A first determination module is configured to extract information of the multi-source fusion data based on an edge server and determine a security level and a delay of the extraction process; The second determining module is configured to determine a target control instruction of the power distribution network system based on the security level and the time delay.

8. A non-volatile storage medium, characterized in that: The non-volatile storage medium includes a stored program, wherein when the program is running, the device where the non-volatile storage medium is located is controlled to execute the control instruction generation method for the power distribution network according to any one of claims 1 to 6.

9. A computer device, characterized in that: include: memory and processor, The memory stores a computer program; The processor is configured to execute a computer program stored in the memory, and when the computer program is run, the processor is enabled to execute the control instruction generation method for a power distribution network according to any one of claims 1 to 6.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the control instruction generation method for a power distribution network according to any one of claims 1 to 6 is implemented.

Images