Information and physical fusion virtual power plant adjustable capacity correction method and system

Abstract

The present application belongs to the technical field of power system reliability evaluation, and provides a virtual power plant adjustable capacity correction method and system considering information and physical fusion, obtains a comprehensive trust degree sequence and attack / abnormal effectiveness indication quantity; the information process is divided into a data acquisition chain, a dispatching control chain and a market transaction chain for the business closed loop of the virtual power plant, an agent trust state sequence reflecting the differences of different chains is constructed; a cross-layer injection mechanism is established, the agent trust state sequence is mapped into two types of quantities, namely control reachable gate control and capacity reduction coefficient, the reachable gate control and the capacity reduction coefficient are synthesized into an injection effect, the up-regulation and down-regulation adjustment capacity boundaries of the virtual power plant are corrected by time-by-time contraction, and the effective adjustment capacity boundary considering the influence of information disturbance is obtained. The present application converts the information side state into time-varying contraction correction of the adjustment capacity boundary, and forms a promised and executable effective adjustment capacity boundary.

Description

Technical Field

[0001] This invention belongs to the field of power system reliability assessment technology, specifically relating to a method and system for correcting the adjustability of a virtual power plant that considers the integration of information and physical systems. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] The operation of virtual power plants relies heavily on information and communication systems to complete closed-loop processes such as measurement and data acquisition, status uploading, scheduling decisions, command issuance, and terminal execution feedback. The boundary of their regulation capacity is not only affected by physical resource constraints, but also significantly constrained by the reliability of information data and control accessibility.

[0004] In actual operation, communication disturbances, link failures, data anomalies, or network attacks can cause information credibility to decay and recover over time. This can lead to theoretically physically feasible adjustment instructions becoming unreachable, delayed, or inconsistently executed at the execution level, resulting in discrepancies between promised capabilities and actual callable capabilities and triggering performance risks. Especially in the multi-stage business chain of a virtual power plant, the attack surface, propagation path, and recovery characteristics of different stages vary, leading to a differentiated dynamic evolution of the credibility state. However, in engineering, it is often difficult to directly measure the true credibility of each stage, and there is a lack of an achievable and reproducible proxy credibility construction method to support subsequent capability boundary adjustments.

[0005] Existing methods for assessing and defining capabilities often assume the reliability of the information-side link, typically only providing a feasible physical boundary or making static margin reservations on top of it. This makes it difficult to explicitly quantify the chain effect of "decreased trustworthiness – unreachable control – shrinking callable capabilities" as the time-varying magnitude of the boundary shrinkage. Even if some methods can identify anomalies or attack risks, they often remain at the alert level, lacking a mechanism to convert comprehensive trust levels and attack / anomaly indicators into calculable injection coefficients. Therefore, they cannot form an effective capability boundary that can be directly used for commitment and invocation under conditions of information disturbance. Summary of the Invention

[0006] To address the aforementioned problems, this invention proposes a method and system for correcting the adjustable capacity of a virtual power plant that considers the integration of information and physical systems.

[0007] According to some embodiments, the present invention adopts the following technical solution:

[0008] A method for adjusting the adjustable capacity of a virtual power plant that considers the integration of information and physical systems includes the following steps:

[0009] Obtain the comprehensive trust level sequence and attack / anomaly validity indicators;

[0010] For the closed-loop business of virtual power plants, the information process is divided into a data acquisition chain, a scheduling and control chain, and a market transaction chain, and a proxy trusted state sequence reflecting the differences of different chains is constructed.

[0011] A cross-layer injection mechanism is established, which maps the agent's trusted state sequence into two types of quantities: control reachability gating and capability reduction coefficient. Control reachability gating is used to characterize whether scheduling instructions can be reached and executed by the terminal, as well as the soft handover process from unreachable to reachable. Capability reduction coefficient is used to characterize the continuous shrinkage of callable capabilities caused by the decrease in trustworthiness.

[0012] The reachability gate and capacity reduction coefficient are combined into an injection effect, and the upward and downward adjustment capacity boundaries of the virtual power plant are corrected hourly to obtain an effective adjustment capacity boundary that takes into account the impact of information disturbances.

[0013] As an alternative implementation method, the process of obtaining the comprehensive trust sequence and attack / anomaly effectiveness indicator includes: at discrete time... t Obtain overall trust level and effective attack / anomaly indicators:

[0014] ;

[0015] ;

[0016] in To assess overall trustworthiness, a higher numerical value indicates better information availability, completeness, authenticity, and reachability. A ( t )=1 indicates that at time 1 t A valid attack exists.

[0017] As an alternative implementation method, the process of dividing the information process into a data acquisition chain, a scheduling and control chain, and a market transaction chain for the virtual power plant business closed loop, and constructing a proxy trusted state sequence reflecting the differences between the different chains includes:

[0018] Considering that it is often difficult to directly measure the true reliability of each link in an engineering process, a link-level proxy reliability vector is constructed to characterize the reliability status of three types of business chains:

[0019] ;

[0020] The trustworthiness of the data acquisition chain proxy is:

[0021] ;

[0022] The trustworthiness of the scheduling control chain agent is:

[0023] ;

[0024] in ≥0 represents the additional sensitivity coefficient of the control chain. The larger the chain of control, the more significant the decrease in its credibility, and the more fragile the chain of control.

[0025] The credibility of the market transaction chain agent is:

[0026] ;

[0027] ;

[0028] in α ∈(0,1] is the smoothing coefficient. α The smaller the value, the smoother the result; the slower the change in credibility due to a decrease in trust level, the more saturated the clip() operator becomes.

[0029] As an alternative implementation, establishing a cross-layer injection mechanism to map the agent trusted state sequence to a control reachability gating process includes:

[0030] The reachable gate is:

[0031] ;

[0032] The change in regulatory capacity resulting from confidence levels above a set threshold is controlled by soft gating, i.e., linear saturation.

[0033] ;

[0034] in , ∈[0,1], As a minimum capability coefficient, The system's adjustable capability when the trust level is 1.

[0035] As an alternative implementation, establishing a cross-layer injection mechanism to map the agent trusted state sequence to a capability reduction factor includes: performing linear normalization above a threshold and normalizing to 0 below the threshold.

[0036] ;

[0037] in To achieve overall trust saturation:

[0038] ;

[0039] Define the prevention and elimination of zero items:

[0040] ;

[0041] Under continuous contraction mode, the hourly reduction factor is:

[0042] ;

[0043] In gated mode:

[0044] ;

[0045] Take the upper / lower adjustment to be consistent:

[0046] ;

[0047] Reduction factor κ ( t Directly from comprehensive trust T ( t Generates and uses it to shrink the up / down adjustment capability boundary, with a threshold of , minimum discount: k min∈[0,1], sensitivity parameter: a >0, b >0, a+b>1 indicates a risk-averse strategy that prioritizes system security; a+b<1 indicates a risk-aggressive strategy that allows for the retention of more capabilities under low trust levels.

[0048] As an alternative implementation, the process of combining the reachability gating and the capability reduction factor into an injection action includes: when simultaneously enabling control reachability gating... When this is done, it is incorporated into the reduction factor to obtain the final injection factor:

[0049] ;

[0050] .

[0051] As an alternative implementation, the process of combining the reachability gate and capacity reduction coefficient into an injection effect, and performing hourly contraction correction on the virtual power plant's upward and downward regulation capacity boundaries, includes:

[0052] By obtaining the final injection coefficient The physical up-and-down adjustment capabilities of a virtual power plant are narrowed down to effective adjustment capabilities that can be promised / called under information disturbances;

[0053] Applying the information-side injection coefficients to the physical adjustability yields the effective boundary:

[0054] ;

[0055] ;

[0056] in, p This indicates a probability guarantee that the adjustment capability will meet the demand, i.e., within a certain percentage.p In this scenario, based on recent scheduling and assessment, a certain generating unit is capable of achieving this adjustment capacity. quantiles p Next scene s Mid-moment t Overall upward adjustment capability quantiles p Next scene s Mid-moment t The overall downward adjustment capability.

[0057] A virtual power plant adjustable capacity correction system considering information and physical integration includes:

[0058] The data acquisition module is configured to acquire a comprehensive trust sequence and an attack / anomaly validity indicator.

[0059] The trusted state construction module is configured to divide the information process into a data acquisition chain, a scheduling and control chain, and a market transaction chain for the virtual power plant business closed loop, and construct a proxy trusted state sequence that reflects the differences between the different chains.

[0060] The mapping module is configured to establish a cross-layer injection mechanism, mapping the agent's trusted state sequence into two types of quantities: control reachability gating and capability reduction coefficient. Control reachability gating is used to characterize whether scheduling instructions can be reached and executed by the terminal, as well as the soft handover process from unreachable to reachable. Capability reduction coefficient is used to characterize the continuous shrinkage of callable capabilities caused by the decrease in trustworthiness.

[0061] The boundary correction module is configured to combine the reachability gate and the capacity reduction coefficient into an injection action to perform hourly contraction correction on the virtual power plant's upward and downward adjustment capacity boundaries, thereby obtaining an effective adjustment capacity boundary that takes into account the impact of information disturbances.

[0062] A computer-readable storage medium for storing computer instructions, which, when executed by a processor, perform the steps in the above method.

[0063] An electronic device includes a memory and a processor, as well as computer instructions stored in the memory and running on the processor, wherein the computer instructions, when executed by the processor, perform the steps in the method described above.

[0064] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0065] This invention can quantify information-side risks such as communication disturbances and network attacks into time-varying contractions of adjustable capability boundaries through a "controllability accessibility gating - capability reduction - injection coefficient" approach, avoiding systemic capability overestimation and performance default risks caused by assuming complete reliability of the information side. It can identify cross-layer risk windows that are "physically feasible but informationally inaccessible / uncontrollable" during dynamic changes in credibility, and adapt to different risk preferences and management strategies in a configurable manner, providing interpretable and calculable quantitative basis for virtual power plant market application and operation risk management.

[0066] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0067] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0068] Figure 1 This is a diagram illustrating the construction of an information and physics-based adjustment curve library operating framework provided in one embodiment.

[0069] Figure 2 This is a schematic diagram illustrating the trust rating and attack event distribution of a virtual power plant in a small-scale scenario provided in one embodiment.

[0070] Figure 3 This is a schematic diagram illustrating the changes in sensing latency, control latency, connection latency, and control deviation in a small-scale scenario provided in one embodiment;

[0071] Figure 4 This is a schematic diagram comparing the physical adjustment capability under a small-scale scenario with the effective adjustment capability under information-side disturbances, provided in one embodiment.

[0072] Figure 5 This is a schematic diagram illustrating the trust rating and attack event distribution of a virtual power plant in a small-scale scenario provided in one embodiment.

[0073] Figure 6 This is a schematic diagram illustrating the changes in sensing latency, control latency, connection latency, and control deviation in a small-scale scenario provided in one embodiment;

[0074] Figure 7 This is a schematic diagram comparing the physical adjustment capability under small-scale scenarios with the effective adjustment capability under information-side disturbances, provided in one embodiment.

[0075] Figure 8 This is a schematic diagram illustrating the trust rating and attack event distribution of a virtual power plant in a medium-scale scenario provided in one embodiment.

[0076] Figure 9 This is a schematic diagram illustrating the changes in sensing latency, control latency, connection latency, and control deviation in a medium-scale scenario provided in one embodiment;

[0077] Figure 10 This is a schematic diagram comparing the physical adjustment capability under medium-scale scenarios and the effective adjustment capability under information-side disturbances, provided in one embodiment.

[0078] Figure 11 This is a schematic diagram illustrating the correspondence between the trust level of a virtual power plant and attack events in a small-scale scenario provided in one embodiment.

[0079] Figure 12 This is a schematic diagram comparing the basic adjustment capability and the reduced effective adjustment capability in a small-scale scenario provided in one embodiment;

[0080] Figure 13 This is a schematic diagram illustrating the decomposition of effective adjustment capability in a small-scale scenario provided in one embodiment.

[0081] Figure 14 This is a schematic diagram illustrating the decomposition of effective adjustment capability in a small-scale scenario provided in one embodiment.

[0082] Figure 15 This is a schematic diagram illustrating the correspondence between the trust level of a virtual power plant and attack events in a small-scale scenario provided in one embodiment.

[0083] Figure 16 This is a schematic diagram comparing the basic adjustment capability and the reduced effective adjustment capability in a small-scale scenario provided in one embodiment;

[0084] Figure 17 This is a schematic diagram illustrating the decomposition of effective adjustment capability in a small-scale scenario provided in one embodiment.

[0085] Figure 18 This is a schematic diagram illustrating the decomposition of effective adjustment capability in a small-scale scenario provided in one embodiment.

[0086] Figure 19 This is a schematic diagram illustrating the changes in trust levels in a virtual power plant under a medium-scale scenario provided in one embodiment.

[0087] Figure 20 This is a schematic diagram comparing the basic adjustment capability and the reduced effective adjustment capability in a medium-scale scenario provided in one embodiment.

[0088] Figure 21 This is a schematic diagram illustrating the decomposition of effective adjustment capability in a medium-scale scenario provided in one embodiment.

[0089] Figure 22This is a schematic diagram illustrating the decomposition of effective adjustment capability in a medium-scale scenario provided in one embodiment.

[0090] Figure 23 This is a schematic diagram illustrating the change in the reliability of the adjustment capability provided in one embodiment. Detailed Implementation

[0091] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0092] It should be noted that the following detailed description is illustrative and intended to provide further explanation of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0093] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0094] Where there is no conflict, the embodiments and features described in this application may be combined with each other.

[0095] Example 1

[0096] This invention proposes a method for information and physical cross-layer injection and regulation capability boundary for virtual power plants. Addressing the common problems in existing regulation capability assessment and commitment boundary setting processes, which often involve "assuming information-side link reliability and making commitments solely based on physical-side boundaries," leading to overestimation of capabilities and performance risks, this invention proposes a computable rule that explicitly maps the trusted state of the information side to a time-varying shrinkage of the regulation capability boundary. This transforms the regulation capability of the virtual power plant from "physically feasible" to an effective boundary that is "informally reachable and executable."

[0097] The core of this invention lies in establishing a "gating-reduction-synthesis injection" mechanism driven by the trusted state of the information side, and applying it to the existing upward / downward adjustment capability curves. Specifically, the information side outputs a comprehensive trust degree sequence and an attack / anomaly effectiveness indicator at discrete moments. Considering the difficulty in directly measuring the true trust degree of each business link in engineering, this invention further constructs link-level proxy trust degree time series for the data acquisition chain, scheduling control chain, and market transaction chain to reflect the differences in attack sensitivity and recovery lag of different chains. Based on this, the comprehensive trust degree or proxy trust degree is mapped to two quantities: a control reachability gating coefficient and a capability reduction coefficient. The control reachability gating is used to characterize the continuous change in the reachability and executability of instructions, and the capability reduction coefficient is used to characterize the continuous contraction of callable capabilities caused by the decrease in trust degree. The two are then synthesized into an injection coefficient.

[0098] By using the aforementioned injection coefficients, this invention performs hourly contraction correction on the virtual power plant's upward and / or downward adjustment capacity curves to obtain an effective adjustment capacity boundary that takes into account the impact of information disturbances. This boundary supports capacity application, reserve commitments, and operational deployment decisions. The contracted upward / downward adjustment capacity boundary can be obtained through deterministic calculations, historical statistics, or other physical assessment methods. This invention does not limit the method of obtaining this boundary, thereby ensuring that this cross-layer injection and contraction mechanism is universally adaptable under different resource structures, different boundary generation methods, and different information security situations.

[0099] The following is a detailed introduction:

[0100] When virtual power plants aggregate resources such as distributed power sources, energy storage, and controllable loads to participate in power energy and reserve operations, their promised regulation capacity is not only affected by the physical constraints of resources, but also significantly depends on the closed-loop reachability and data reliability of the information side's "measurement acquisition—communication transmission—scheduling decision—command issuance—terminal execution". Traditional methods often assume the reliability of the information side and only give the promised capacity based on the feasible boundary of the physical side. This leads to a window of "physical feasibility but information unreachability / uncontrollability" under conditions of communication disturbances, data anomalies, or network attacks, thereby overestimating the callable capacity and amplifying the performance risk. To address this, this invention proposes a cross-layer injection and boundary contraction method based on the trusted state of the information side. This method transforms the information side state into a time-varying contraction correction of the regulation capacity boundary, forming a promised and executable effective regulation capacity boundary.

[0101] The method of the present invention includes the following process: First, a comprehensive trust sequence and attack / anomaly validity indication are obtained on the information side to describe the impact of communication disturbances, data anomalies and network attacks on information availability and control accessibility; and for the virtual power plant business closed loop, the information process is divided into a data acquisition chain, a scheduling control chain and a market transaction chain, and a proxy trust state sequence reflecting the differences of different chains is constructed to adapt to the situation in engineering where it is difficult to directly measure the true trustworthiness of each link.

[0102] Subsequently, a cross-layer injection mechanism is established based on the trusted state, mapping the trusted state on the information side to two types of actions: "control reachability gating" and "capacity reduction." Control reachability gating characterizes whether scheduling commands can reach and be executed by the terminal, as well as the soft handover process from unreachable to reachable. Capacity reduction characterizes the continuous contraction of callable capacity caused by a decrease in trustworthiness. By combining gating and reduction into an injection action, the upward and downward adjustment capacity boundaries (or adjustment capacity curves) of the virtual power plant are corrected hourly to obtain an effective adjustment capacity boundary that takes into account the impact of information disturbances.

[0103] The aforementioned adjustment / reduction of the regulatory capability boundary can be obtained through deterministic calculation, historical statistics, or other physical-side assessment methods; this invention does not limit the method of acquisition. Through the above methods, this invention can explicitly quantify information-side risks as a time-varying contraction of the regulatory capability boundary in the form of "gating-reduction-injection," avoiding a systematic overestimation of the committable capability under network disturbances or attacks, and providing an interpretable basis for identifying critical risk periods and weak points.

[0104] The specific steps include:

[0105] Step 1: Output of Layered Trust Chains and Credibility Time Series

[0106] This section addresses the information loop of a virtual power plant, encompassing "measurement acquisition—communication transmission—dispatch control—market interaction," considering the differentiated attenuation and recovery characteristics of network disturbances, data anomalies, and attacks across different business chains. To ensure feasibility and reproducibility, this patent adopts a modeling approach of "comprehensive trust output → link-level proxy trust decomposition → cross-layer injection": the information layer first outputs comprehensive trust and effective attack indications at discrete moments, then constructs proxy trust time series for the data acquisition chain, dispatch control chain, and market transaction chain based on these, providing a unified input for subsequent "gating—reduction—curve library injection." Physical-side dispatch modeling is more extensive and will not be elaborated here; the detailed process is as follows... Figure 1 As shown.

[0107] Step 1.1 Modeling the Integration of Trust Level and Attack Indicators

[0108] Information layer at discrete time t(Updated hourly) Outputs overall trust level and number of valid attack / anomaly indicators:

[0109] (1)

[0110] (2)

[0111] in To assess overall trustworthiness, a higher value indicates better information availability, completeness, authenticity, and reachability. A ( t )=1 indicates that at time 1 t A valid attack exists.

[0112] Step 1.2 Constructing the trust vector of the link-level agent (trust decomposition based on the trust chain)

[0113] Considering that it is often difficult to directly measure the true reliability of each link in an engineering process, this paper constructs a link-level agent reliability vector to characterize the reliability status of three types of business chains:

[0114] (3)

[0115] Trustworthiness of the data acquisition chain proxy (equivalent to overall trust level)

[0116] (4)

[0117] Trustworthiness of the scheduling control chain proxy (more sensitive to effective attacks)

[0118] (5)

[0119] in ≥0 represents the additional sensitivity coefficient of the control chain. The larger the chain of control, the more significant the decrease in its credibility, and the more fragile the chain of control becomes.

[0120] Market transaction chain agent credibility (first-order low-pass / exponential smoothing, reflecting "passivation / lag")

[0121] (6)

[0122] (7)

[0123] in α ∈(0,1] is the smoothing coefficient. α The smaller the value, the smoother the result, and the slower the change in credibility due to a decrease in trust level. clip() is the saturation truncation operator.

[0124] Step 2: Injecting Trust Level into the Curve Library: Modeling Gating and Reduction Mechanisms

[0125] This section presents cross-layer injection rules: mapping the overall trust level (or link-level credibility) to two quantities affecting the boundary of regulatory capability: control accessibility gating and capability reduction coefficient. These two factors together characterize whether "physically feasible capability" can be committed to and invoked under information disturbance conditions, and provide injection coefficients for subsequently shrinking the physical quantile curve library into an effective curve library.

[0126] Step 2.1 Control reachability gating (soft gating modeling)

[0127] Define reachability gates:

[0128] (8)

[0129] To avoid discrete jumps, the change in regulation capability caused by confidence levels above a certain threshold is controlled by soft gating (linear saturation form):

[0130] (9)

[0131] in , ∈[0,1]. The safety margin coefficient represents the system's ability to adjust even when network communication is completely unreliable. This represents the system's adjustable capability when the trust level is equal to 1.

[0132] Step 2.2 Capacity Reduction Factor: Threshold Normalization and Sensitivity Strategy

[0133] Reduction factor κ ( t Directly from comprehensive trust T ( t This is generated and used to shrink the up / down adjustment capability boundary. Threshold: , minimum discount: k min∈[0,1], sensitivity parameter: a >0, b >0. a+b>1 (convex function): represents a risk-averse strategy, where a slight decrease in trust leads to a significant contraction in regulation capacity, prioritizing system safety; a+b<1 (concave function): represents a risk-aggressive strategy, allowing more capacity to be retained even at low trust levels. In this paper, a=1 and b=1 are set. The specific parameter tuning can be optimized through game theory based on the default penalty weights in the electricity market and the operating risk costs of the generating units.

[0134] First, perform linear normalization above the threshold (and normalize to 0 below the threshold):

[0135] (10)

[0136] in To achieve overall trust saturation:

[0137] (11)

[0138] Define the prevention and elimination of zero items:

[0139] (12)

[0140] Under continuous contraction mode, the hourly reduction factor is:

[0141] (13)

[0142] In gated mode:

[0143] (14)

[0144] Take the upper / lower adjustment to be consistent:

[0145] (15)

[0146] Step 2.3 Combination of gating and reduction

[0147] When both control reachability gates are enabled When this is done, it is incorporated into the reduction factor to obtain the final injection factor:

[0148] (16)

[0149] (17)

[0150] The process of combining the reachability gate and capacity reduction coefficient into an injection effect, and performing hourly contraction correction on the virtual power plant's upward and downward regulation capacity boundaries, includes:

[0151] By obtaining the final injection coefficient The physical up-and-down adjustment capabilities of a virtual power plant are narrowed down to effective adjustment capabilities that can be promised / called under information disturbances;

[0152] Applying the information-side injection coefficients to the physical adjustability yields the effective boundary:

[0153]

[0154]

[0155] in, p This indicates a probability guarantee that the adjustment capacity will meet the demand, i.e., within a certain percentage. p In this scenario, based on recent scheduling and assessment, a certain generating unit is capable of achieving this adjustment capacity. quantiles p Next scenes Mid-moment t Overall upward adjustment capability quantiles p Next scene s Mid-moment t The overall downward adjustment capability.

[0156] To verify the effectiveness of the method provided in this embodiment, a simulation environment was built based on the Python 3.12 platform. A virtual power plant simulation control system covering day-ahead, intraday, and real-time scenarios was designed. The system was tested using an IEEE 30-node network, and the Gurobi solver was used for optimized scheduling. Day-ahead scheduling used a one-hour time step, intraday scheduling used a 15-minute time step, and real-time scheduling used a 15-second time step.

[0157] To verify the robustness of the proposed strategy under network threats, this paper constructs a NASim-based mirror communication network layer on top of the physical topology and designs a dynamic attack mechanism incorporating four typical behaviors: brute-force attacks and intensive scanning are used to simulate high-frequency indiscriminate probing and denial-of-service risks; adaptive attacks utilize a feedback mechanism to dynamically adjust the intrusion strategy based on the defense status; and advanced persistent threats simulate covert long-term attacks using a "reconnaissance-penetration-lateral movement" kill chain pattern. These attack sequences are randomly injected into the scheduling period following a Poisson distribution, generating multi-dimensional heterogeneous data including response latency fluctuations, vulnerability trigger records, and service interruption status, providing a training environment for feature extraction and real-time evaluation of the dynamic trust model.

[0158] To address the uncertainty of the source load, a stacked autoencoder (SAE) was used for scene generation and feature extraction, and the Sobol method was employed for parameter sensitivity analysis to ensure the rationality of the model parameter settings and the reproducibility of the experimental results. The key parameter settings of the simulation environment are shown in Table 1.

[0159] Table 1 Simulation Environment Parameter Configuration

[0160]

[0161] To generate wind and solar load scenarios that conform to statistical regularity and preserve temporal continuity, this paper adopts a scenario generation method based on empirical Copula functions and stacked autoencoders (SAEs) to model the uncertainties of wind power, photovoltaics, and loads.

[0162] To verify the scalability and stability of the proposed method under different system scales, three test systems of different scales—tiny, small, and medium—were set up for comparison. The tiny system consisted of six generator nodes (1, 2, 5, 8, 11, and 13) from the IEEE 30-node system; the small system added some critical load nodes to this system; and the medium system used the complete IEEE 30-node system. Meanwhile, the attack coverage was gradually increased with the system scale to evaluate the changes in the adjustment capability assessment and reliability calculation results under different attack surfaces and network coupling strengths.

[0163] The impact of information layer trust evolution and attack triggering on trusted injection

[0164] (1) Trust time-varying trajectory and anti-interference characteristics in multiple scenarios

[0165] like Figure 2 , Figure 5 and Figure 8 As shown, the overall trust level exhibits significantly differentiated evolution characteristics under different system configurations (Tiny / Small / Medium). In the Tiny scenario, which has weaker anti-interference capabilities, the trust score rapidly drops below the security threshold in the early stages of an attack and remains in a low-level oscillation for a long period, accompanied by high-frequency "attack successful" markers, indicating that the system is in a high-risk state. The Small scenario shows intermittent fluctuations in trust level, with a sudden drop only at specific moments, reflecting the existence of sparse vulnerability windows in the system. In contrast, the Medium scenario demonstrates strong robustness, with the trust level quickly converging and stabilizing at a high level after a brief fluctuation, indicating that the information layer under this configuration has a strong steady-state trust level and can effectively suppress attack disturbances.

[0166] (2) Trust-based gating and reduction mapping mechanism

[0167] Figure 3 , Figure 6 , Figure 9 This study reveals a dynamic mapping mechanism from "information trust" to "physical constraints." Observations show that the gating coefficient and reduction coefficient are highly sensitive to changes in trust level: in the low-trust phase (such as the first 200 hours of the Tiny scenario), due to extremely low trust, the gating coefficient is directly cut off or deeply reduced, at which point cross-layer injection exhibits strong suppression of adjustment capabilities; as trust gradually recovers and exceeds the threshold, the reduction coefficient tends to saturate, and the system smoothly transitions to a weakly suppressed or lossless state. This nonlinear mapping mechanism ensures that physical resources can be forcibly locked or devalued when information is unreliable, preventing the execution of untrustworthy instructions.

[0168] (3) Quantitative assessment of effective regulation capacity under information disturbance

[0169] Figure 4 , Figure 7 and Figure 10 The diagram visually demonstrates the effect of information-physical coupling on correcting the final regulatory capability boundary. The green dashed line represents the ideal quantile capability considering only physical constraints (P95), while the blue solid line represents the effective capability boundary after incorporating information perturbations. It can be seen that during periods of decreased trust due to cyberattacks, the effective capability boundary shrinks significantly. Particularly in the Tiny scenario, numerous extreme cases emerge where physical capabilities are abundant but effective capabilities are zero, confirming that relying solely on physical assessments for market commitments carries a serious risk of systemic overestimation. The proposed fusion method explicitly quantifies this risk bias, providing a credible scheduling boundary that satisfies information security constraints.

[0170] Adjustment capacity and resource contribution vary with trust level

[0171] (1) Differences in regulatory capacity and scale dependence caused by the integration of information and physical systems

[0172] Figure 11 , Figure 12 , Figure 15 , Figure 16 , Figure 19 and Figure 20 The diagram visually illustrates the differentiated response of VPP's adjustment capability to information disturbances of equal intensity under three different scale scenarios. The dashed line in the figure represents the basic quantile curve considering only physical constraints, while the solid line represents the effective capability boundary after incorporating information trust. The difference between the two represents the capability reduction caused by information risk.

[0173] The comparison revealed that the impact of information disturbances on performance risk has a significant scale dependence: the Tiny scenario exhibits extremely high vulnerability, with effective up / down capabilities rapidly collapsing to zero in the early stages of the attack, a stark contrast to the still high theoretical physical value, indicating that small-scale aggregates lack information-side buffering and recovery mechanisms; the Small scenario exhibits significant "pulsating" vulnerability characteristics, with effective capabilities experiencing intermittent interruptions and recoveries as trust levels fluctuate, accurately capturing performance risks during the attack window; the Medium scenario, on the other hand, demonstrates strong resilience, with the reduction factor only slightly decreasing at the moment of the attack, and the effective capability boundary basically conforming to the physical boundary, indicating that large-scale systems can effectively mitigate information-side disturbances through resource sharing and a high level of basic trust.

[0174] (2) Dynamic restructuring of resource structure and risk period positioning

[0175] Figure 13 , Figure 14 , Figure 17 , Figure 18 , Figure 21 and Figure 22The decomposition stacked diagram of regulatory capacity further reveals the changes in the composition of effective capacity from a micro perspective. The cross-layer injection mechanism not only compresses the total boundary at the system level, but also triggers a dynamic reconstruction of the "callable resource structure".

[0176] In the Tiny scenario, the large areas of blank space in the stacked graph intuitively reflect the synchronous failure of all sub-resources under low trust conditions. In the Small scenario, the stacked graph clearly depicts the temporal distribution of capability gaps, meaning that the effective contributions of corresponding units (GT, ESS, etc.) are forcibly blocked during specific attack periods (such as steps 25 and 45). This demonstrates that the proposed method not only prevents the system from "capability overestimation" and "market default" due to blind reliance on physical predictions, but also accurately identifies key risk periods and weak links through visualization, providing an executable security boundary for VPP's multi-confidence level declaration and rolling scheduling in information-disrupted environments.

[0177] Reliability comparison between physical curve libraries and fusion curve libraries

[0178] Figure 23 The power supply reliability comparison is presented in the medium scenario, with the statistics in hourly increments. The shaded area represents the range of P50–P99, and the broken line represents the median.

[0179] The graph shows that the median reliability of PHY was higher than that of CPL for most periods, and the PHY range shifted upwards overall. This is because PHY does not account for information-side reduction, which is equivalent to using a more optimistic callable capacity boundary in scheduling verification; while CPL maps communication disturbances and observability / controllability degradation to effective capacity contraction, thus giving a more conservative but more realistic reliability estimate across layers.

[0180] During certain periods (e.g., from morning to around noon), CPL reliability experienced a significant trough, while PHY remained at a relatively high level. This indicates that these periods present a typical cross-layer risk window where "physical capabilities appear sufficient, but information capabilities lead to unavailability": that is, physical capabilities are not insufficient, but due to decreased reachability of trusted state / control, the committable boundary shrinks, causing a decrease in scheduling success rate.

[0181] Observing the shaded bandwidths of P50–P99 in the graph, it is evident that the bandwidth of CPL exhibits a significant expansion and contraction during high-risk periods (e.g., 11:00–13:00). This indicates that the source of uncertainty has expanded from a single physical randomness to a dual superposition of "physical + information," significantly increasing the dispersion of the system's operational boundaries. This means that if scheduling decision-makers require a higher level of confidence (e.g., P99), a larger safety margin needs to be reserved in CPL mode, resulting in a substantial reduction in the lower limit of the reliability assessment value. This bandwidth change quantifies the "confidence cost" brought about by information disturbances; that is, in an information-insecure environment, to obtain a high-confidence commitment, one must accept a lower reliability expectation.

[0182] Example 2

[0183] A virtual power plant adjustable capacity correction system considering the integration of information and physical systems includes:

[0184] The data acquisition module is configured to acquire a comprehensive trust sequence and an attack / anomaly validity indicator.

[0185] The trusted state construction module is configured to divide the information process into a data acquisition chain, a scheduling and control chain, and a market transaction chain for the virtual power plant business closed loop, and construct a proxy trusted state sequence that reflects the differences between the different chains.

[0186] The mapping module is configured to establish a cross-layer injection mechanism, mapping the agent's trusted state sequence into two types of quantities: control reachability gating and capability reduction coefficient. Control reachability gating is used to characterize whether scheduling instructions can be reached and executed by the terminal, as well as the soft handover process from unreachable to reachable. Capability reduction coefficient is used to characterize the continuous shrinkage of callable capabilities caused by the decrease in trustworthiness.

[0187] The boundary correction module is configured to combine the reachability gate and the capacity reduction coefficient into an injection action to perform hourly contraction correction on the virtual power plant's upward and downward adjustment capacity boundaries, thereby obtaining an effective adjustment capacity boundary that takes into account the impact of information disturbances.

[0188] Example 3

[0189] A computer-readable storage medium for storing computer instructions, which, when executed by a processor, perform the steps in the above method.

[0190] Example 4

[0191] An electronic device includes a memory and a processor, as well as computer instructions stored in the memory and running on the processor, wherein the computer instructions, when executed by the processor, perform the steps in the method described above.

[0192] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of one or more computer-usable storage media (including, but not limited to, disk storage, etc.) containing computer-usable program code. CD - ROM It takes the form of a computer program product implemented on (such as optical memory, etc.).

[0193] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0194] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0195] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0196] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made by those skilled in the art without creative effort within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for adjusting the adjustable capacity of a virtual power plant considering the integration of information and physical systems, characterized in that, Includes the following steps: Obtain the comprehensive trust level sequence and attack / anomaly validity indicators; For the closed-loop business of virtual power plants, the information process is divided into a data acquisition chain, a scheduling and control chain, and a market transaction chain, and a proxy trusted state sequence reflecting the differences of different chains is constructed. A cross-layer injection mechanism is established, which maps the agent's trusted state sequence into two types of quantities: control reachability gating and capability reduction coefficient. Control reachability gating is used to characterize whether scheduling instructions can be reached and executed by the terminal, as well as the soft handover process from unreachable to reachable. Capability reduction coefficient is used to characterize the continuous shrinkage of callable capabilities caused by the decrease in trustworthiness. The reachability gate and capacity reduction coefficient are combined into an injection effect, and the upward and downward adjustment capacity boundaries of the virtual power plant are corrected by time-by-time contraction to obtain an effective adjustment capacity boundary that takes into account the impact of information disturbance. The process of combining the reachability gating and capability reduction factor into an injection action includes: when simultaneously enabling control reachability gating... When this is done, it is incorporated into the reduction factor to obtain the final injection factor: ； ； To control reachability gating, , They represent the reduction factors, respectively. The upper / lower adjustment capability boundary; The process of combining the reachability gate and capacity reduction coefficient into an injection effect, and performing hourly contraction correction on the virtual power plant's upward and downward regulation capacity boundaries, includes: By obtaining the final injection coefficient The physical up-and-down adjustment capabilities of a virtual power plant are narrowed down to effective adjustment capabilities that can be promised / called under information disturbances; Applying the information-side injection coefficients to the physical adjustability yields the effective boundary: ； ； in, p This indicates a probability guarantee that the adjustment capacity will meet the demand, i.e., within a certain percentage. p In this scenario, based on recent scheduling assessments, a certain generating unit is capable of achieving this adjustment capacity. quantiles p Next scene s Mid-moment t Overall upward adjustment capability quantiles p Next scene s Mid-moment t The overall downward adjustment capability.

2. The virtual power plant adjustable capacity correction method considering information and physical integration as described in claim 1, characterized in that, The process of obtaining the comprehensive trust sequence and attack / anomaly effectiveness indicators includes: at discrete time... t Obtain overall trust level and effective attack / anomaly indicators: in To assess overall trustworthiness, a higher numerical value indicates better information availability, completeness, authenticity, and reachability. A ( t )=1 indicates that at time 1 t A valid attack exists.

3. The virtual power plant adjustable capacity correction method considering information and physical integration as described in claim 1, characterized in that, The process of dividing the information process into a data acquisition chain, a scheduling and control chain, and a market transaction chain for a virtual power plant business closed loop, and constructing a proxy trusted state sequence reflecting the differences between the different chains includes: Considering that it is often difficult to directly measure the true reliability of each link in an engineering process, a link-level proxy reliability vector is constructed to characterize the reliability status of three types of business chains: The trustworthiness of the data acquisition chain proxy is: The trustworthiness of the scheduling control chain agent is: in ≥0 represents the additional sensitivity coefficient of the control chain. The larger the chain of control, the more significant the decrease in its credibility, and the more fragile the chain of control. The credibility of the market transaction chain agent is: in α ∈(0,1] is the smoothing coefficient. α The smaller the value, the smoother the result; the slower the change in credibility due to a decrease in trust level. `clip()` is a saturation truncation operator. To assess overall trust levels.

4. The virtual power plant adjustable capability correction method considering information and physical integration as described in claim 1, characterized in that, The process of establishing a cross-layer injection mechanism to map the agent's trusted state sequence to a control reachability gate includes: The reachable gate is: The change in regulatory capacity resulting from confidence levels above a set threshold is controlled by soft gating, i.e., linear saturation. in , ∈[0,1], As a minimum capability coefficient, This represents the system's adjustable capability when the trust level is equal to 1. Trustworthiness of the scheduling control chain agent.

5. The virtual power plant adjustable capacity correction method considering information and physical integration as described in claim 1, characterized in that, The process of establishing a cross-layer injection mechanism and mapping the agent's trusted state sequence to a capability reduction factor includes: performing linear normalization above a threshold and normalizing to 0 below the threshold. in To achieve overall trust saturation: Define the prevention and elimination of zero items: Under continuous contraction mode, the hourly reduction factor is: In gated mode: Take the upper / lower adjustment to be consistent: Reduction factor Directly based on overall trust level T ( t Generates and uses it to shrink the up / down adjustment capability boundary, with a threshold of , minimum discount: ∈[0,1], sensitivity parameter: a >0, b >0, a+b>1 indicates a risk-averse strategy that prioritizes system security; a+b<1 indicates a risk-aggressive strategy that allows for the retention of more capabilities under low trust levels.

6. A virtual power plant adjustable capacity correction system considering the integration of information and physical systems, employing the method of claim 1, characterized in that, include: The data acquisition module is configured to acquire a comprehensive trust sequence and an attack / anomaly validity indicator. The trusted state construction module is configured to divide the information process into a data acquisition chain, a scheduling and control chain, and a market transaction chain for the virtual power plant business closed loop, and construct a proxy trusted state sequence that reflects the differences between the different chains. The mapping module is configured to establish a cross-layer injection mechanism, mapping the agent's trusted state sequence into two types of quantities: control reachability gating and capability reduction coefficient. Control reachability gating is used to characterize whether scheduling instructions can be reached and executed by the terminal, as well as the soft handover process from unreachable to reachable. Capability reduction coefficient is used to characterize the continuous shrinkage of callable capabilities caused by the decrease in trustworthiness. The boundary correction module is configured to combine the reachability gate and the capacity reduction coefficient into an injection action to perform hourly contraction correction on the virtual power plant's upward and downward adjustment capacity boundaries, thereby obtaining an effective adjustment capacity boundary that takes into account the impact of information disturbances.

7. A computer-readable storage medium, characterized in that, Used to store computer instructions, which, when executed by a processor, complete the steps of the method according to any one of claims 1-5.

8. An electronic device, characterized in that, It includes a memory and a processor, as well as computer instructions stored in the memory and running on the processor, which, when executed by the processor, perform the steps of the method according to any one of claims 1-5.

Images