Frequency stability control method for accessing new energy to offshore energy platform

Abstract

The invention discloses a frequency stability control method for accessing new energy to an offshore energy platform, and the method comprises the steps: firstly carrying out the state recognition and parameter extraction of frequency modulation resources, building a frequency-power response model of the frequency modulation resources, carrying out the superposition construction of a system frequency response total model, and carrying out the packaging of a dictionary structure; based on a frequency safety index, disturbance occurrence and grade are judged in real time, and a frequency change trend is dynamically estimated; according to the disturbance level and the response characteristics of various resources, a fast-medium-slow layered frequency modulation control strategy is constructed, different types of resources are scheduled in stages for cooperative adjustment, the frequency state is monitored, and control parameters are dynamically corrected; constructing a fuzzy logic controller to process nonlinear disturbance, and introducing a whale optimization algorithm to set and optimize parameters of the fuzzy controller; and finally, dynamically correcting the control strategy according to the frequency control result and the resource utilization rate index. The method can guarantee the frequency stability, reduces the frequency modulation cost, and improves the frequency control precision, the response speed and the overall operation toughness.

Description

Technical Field

[0001] This invention belongs to the technical field of power system frequency control and intelligent dispatch optimization, and mainly relates to a frequency stability control method for new energy access to offshore energy platforms. Background Technology

[0002] With the accelerating global energy structure transformation, offshore renewable energy platforms (such as offshore wind power and tidal power platforms) are gradually becoming an important component of renewable energy. Compared to onshore power grids, offshore energy systems are characterized by independent operation and self-contained nature. Their operating environment is more severe, communication bandwidth is limited, maintenance is more difficult, and the inertial support capacity of the power system is significantly reduced. Against this backdrop, how to achieve efficient frequency control of offshore isolated power systems has become a key technical challenge to ensure their stable operation and the efficiency of renewable energy absorption.

[0003] Traditional frequency control strategies generally rely on thermal power units or hydropower systems with high inertia and regulation capabilities. However, due to space and environmental constraints, most offshore platforms use new energy sources such as wind power and photovoltaics as their main energy source. These sources lack inertia, resulting in strong and unpredictable power output fluctuations, leading to frequent system frequency fluctuations and a high risk of frequency drops or even system collapse. Existing frequency control methods, such as PID controllers based on fixed parameters or simple empirical scheduling strategies, are often ill-suited to this highly volatile and uncertain operating environment. Their main shortcomings are twofold: first, slow response speed and low regulation accuracy, making it difficult to cope with sudden disturbances; second, a lack of unified modeling and coordinated control among multiple heterogeneous resources (such as energy storage, high-voltage direct current, pumped storage, and interruptible loads), resulting in low resource allocation efficiency and high system operating costs.

[0004] Meanwhile, with the development of smart grid technology, the new generation of frequency control systems has gradually integrated advanced methods such as fuzzy logic, adaptive adjustment, and intelligent optimization algorithms. Especially with the increasing proportion of renewable energy, the traditional approach of "centralized control + rigid scheduling" is being replaced by a "layered collaboration + flexible control" mechanism. The new control system not only requires the ability to dynamically sense disturbance levels but also needs to achieve collaborative optimization among resources with different response rates to adapt to complex and ever-changing operating scenarios. Summary of the Invention

[0005] This invention addresses the technical challenges of severe frequency fluctuations, delayed frequency regulation response, and low resource allocation efficiency in isolated power grids. It provides a frequency stability control method for integrating new energy sources into offshore energy platforms. First, the state of frequency regulation resources is identified and parameters are extracted to establish their frequency-power response models. The frequency domain response functions of various frequency regulation resources are transformed into frequency-power transfer functions, and these are superimposed to construct a comprehensive system frequency response model, which is then encapsulated into a dictionary structure. Based on frequency safety indicators, the occurrence and level of disturbances are determined in real time. Using differential algorithms and approximate linear models, the frequency change trend is dynamically estimated, and disturbance levels are classified according to the frequency change trend. Based on the disturbance level and the response characteristics of various resources, a "fast-medium-slow" hierarchical frequency regulation control strategy is constructed, scheduling different types of resources for coordinated adjustment in stages, monitoring frequency status, and dynamically correcting control parameters. A fuzzy logic controller is constructed to handle nonlinear disturbances. Using frequency deviation and rate of change as input, seven types of fuzzy language sets and membership functions are constructed to achieve fuzzy control of frequency regulation power. A whale optimization algorithm is introduced to tune and optimize the parameters of the fuzzy controller. Finally, a real-time evaluation and feedback mechanism is constructed to dynamically correct the control strategy based on the frequency control results and resource utilization indicators. The method of this invention reduces frequency modulation costs while ensuring frequency stability, enhances the frequency control accuracy, response speed and overall operational resilience of the system under complex operating conditions, and comprehensively promotes the intelligent development of frequency control for offshore energy platforms.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is: a frequency stability control method for new energy access to offshore energy platforms, comprising at least the following steps:

[0007] S1: Perform state identification and parameter extraction on frequency regulation resources, and establish their frequency-power response model. The frequency regulation resources include at least battery energy storage systems, wind power / photovoltaic systems with virtual inertia control, high voltage DC equipment, pumped storage systems, and interruptible loads. The frequency domain response functions of various frequency regulation resources are converted into frequency-power transfer function forms, and after superimposing to construct the overall frequency response model of the system, they are encapsulated into a dictionary structure.

[0008] S2: Based on frequency safety indicators, the occurrence and level of disturbances are determined in real time. Based on differential algorithms and approximate linear models, the frequency change trend is dynamically estimated, and the disturbance level is classified according to the frequency change trend. The frequency safety indicators include at least frequency deviation, rate of change, and prediction of the lowest frequency point.

[0009] S3: Based on the disturbance level and the response characteristics of various resources, a "fast-medium-slow" hierarchical frequency modulation control strategy is constructed to coordinate the adjustment of different types of resources in stages, monitor the frequency status and dynamically correct the control parameters;

[0010] S4: Construct a fuzzy logic controller to handle nonlinear disturbances. With frequency deviation and rate of change as input, construct seven types of fuzzy language sets and membership functions to realize frequency modulation power fuzzy control, and introduce the whale optimization algorithm to tune and optimize the parameters of the fuzzy controller.

[0011] S5: Construct a real-time evaluation and feedback mechanism to dynamically adjust the control strategy based on the frequency control results and resource utilization indicators.

[0012] As an improvement of the present invention, step S1 specifically includes the following steps:

[0013] S11: Extract the operating parameters of various frequency regulation resources. The frequency regulation resources include at least battery energy storage systems, wind power / photovoltaic systems with virtual inertia control, high voltage DC equipment, pumped storage systems, and interruptible loads. The operating parameters generate a resource status data table in a unified format, including but not limited to rated capacity, real-time output power, standby regulation capacity, response delay, state of charge, voltage level, and control strategy.

[0014] S12: Construct frequency response models for various resources based on their control characteristics. For battery energy storage systems and high-voltage DC equipment, a linear function of frequency deviation is used for modeling. For wind power / photovoltaic systems with virtual inertia control, a joint model of virtual inertia and droop control is established by combining the rate of frequency change. For pumped storage systems, a first-order hysteresis model in the Laplace domain is used to represent their inertia and delay characteristics.

[0015] S13: Utilizing the linear superposition property of the Laplace transform, the frequency domain response functions of various resources are uniformly transformed into frequency-power transfer function forms, and the system's overall frequency response model is constructed by superposition; the overall frequency response power is specifically:

[0016]

[0017] Among them, G i (s) represents the frequency form of the response model corresponding to the i-th type of resource, ω i Let f(s) be the scheduling weight of the i-th resource, and Δf(s) be the frequency deviation signal.

[0018] S14: Encapsulate the running parameters and transfer functions of the above resources into a dictionary structure for storage. The dictionary structure supports dynamic updates and scalable calls.

[0019] As an improvement to the present invention, in step S12, the battery energy storage system and the high-voltage DC equipment are modeled using a linear function of frequency deviation, specifically as follows:

[0020]

[0021] Among them, △P fast(t) The adjusted output power of the energy storage system at time t, k b For frequency response gain coefficient, SoC min and SoC max These represent the minimum and maximum allowable states of charge of the energy storage system, respectively, Δf(t) = f ref -f(t) represents the system frequency deviation, f ref The reference frequency is set for the system, and f(t) is the real-time frequency measured.

[0022] The wind / photovoltaic system with virtual inertia control establishes a joint model of virtual inertia and droop control based on the frequency change rate, specifically as follows:

[0023]

[0024] Among them, VP VIC (t) represents the output regulation power of the virtual inertial system, k p and k d These represent the frequency droop control coefficient and the virtual inertia control coefficient, respectively. It is the rate of change of frequency deviation;

[0025] The pumped storage system uses a first-order hysteresis model in the Laplace domain to represent its inertia and delay characteristics. The specific first-order delay response model is as follows:

[0026]

[0027] Where s is the complex frequency domain variable under the Laplace transform, and VP i (s) represents the frequency domain output power of the i-th type of frequency modulation resource, Δf(s) represents the frequency deviation signal, and K i T represents the power adjustment caused by a unit change in frequency. i Let τ be the first-order inertial time constant. i To respond to the delay, The transfer function form representing pure delay.

[0028] As another improvement of the present invention, in step S2, the frequency deviation change is detected in real time using a differential algorithm to determine whether a disturbance event has occurred. Specifically, if the static deviation exceeds the limit |Δf(t)|>Δf th RoCoF exceeded limits Or the rate of change in energy power exceeds the fluctuation threshold. This is determined to be a disturbance, triggering emergency frequency control; where △f th R is the frequency deviation trigger threshold. th P is the threshold for the rate of change of frequency. VRE The sum of renewable energy power, where α represents the fluctuation threshold;

[0029] The frequency change rate RoCoF is estimated in real time using median filtering and a differential algorithm, specifically:

[0030]

[0031] Where Δt is the sampling time interval; if the absolute value of RoCoF is greater than R th If the disturbance is high-intensity, then the fast response resources will be prioritized.

[0032] Using a frequency linear variation approximation model, the time t for the occurrence of the lowest frequency point is estimated based on the current RoCoF value. nadir With frequency minimum value f nadir :

[0033]

[0034] f nadir =f ref -RoCoF·t nadir

[0035] Among them, △f safe For the safety margin of frequency offset, The time position of the undamped sine wave at the first extreme point is given. If the predicted result is lower than the system frequency safety limit, emergency frequency regulation resources will be triggered immediately.

[0036] As another improvement of the present invention, the "fast-medium-slow" hierarchical frequency modulation control strategy in step S3 specifically includes the following steps:

[0037] S31: When a medium-to-high intensity disturbance occurs, immediately call up rapid resources, including but not limited to battery energy storage systems and high-voltage DC equipment. Based on their frequency-power response model, output regulating power to achieve initial frequency suppression and rapid stabilization.

[0038] S32: After the rapid resources stabilize, the frequency first-order adjustment stage is entered, and medium-speed resources are called up. The medium-speed resources include, but are not limited to, pumped storage systems. The optimal commissioning time and adjustment range are calculated based on their first-order lag model to support the frequency recovery.

[0039] S33: If the system frequency recovery speed is slow or the early warning frequency adjustment capability is insufficient, slow resources will be called. The slow resources include, but are not limited to, interruptible loads. They will be sorted according to their importance or scheduling level, and loads with low importance will be cut off first to minimize the interference to system operation and production.

[0040] S34: The control system continuously monitors the frequency change trajectory to determine whether the current frequency has recovered to the permissible operating range. If the frequency still does not meet the requirements, the adjustment coefficient (k) will be adjusted according to the actual deviation. bk p and k d ), power adjustment range K i In addition, it can change the load shedding strategy to achieve secondary modification and closed-loop optimization of the control strategy.

[0041] As another improvement of the present invention, in step S33, the load shedding strategy for interruptible loads is represented by the following linear weighted model:

[0042]

[0043] Among them, △P cut Represents the total load power currently being cut off, where n is the total number of loads that can be cut off, and c i δ is the unit power of the i-th load. i It is a load shedding state variable.

[0044] As another improvement of the present invention, step S4 specifically includes the following steps:

[0045] S41: Construct a fuzzy controller whose basic structure includes two input variables and one output variable. The input variables are the frequency deviation Δf and the rate of change of frequency. The output variable is the power adjustment command u(t), and the two input variables are divided into 7 fuzzy language sets: NB (large negative), NM (medium negative), NS (small negative), ZE (zero), PS (small positive), PM (medium positive), PB (large positive), corresponding to 7 triangular membership functions;

[0046] S42: Design a fuzzy inference rule table to map the input language set to the output control action, and realize inference decision-making through the fuzzy rule base; the inference output result is defuzzified using the centroid method to obtain the precise adjustment power value, and output to the fast resource execution control command.

[0047] S43: The optimal parameter combination of the controller is searched using the whale optimization algorithm. The optimal parameter combination includes at least fuzzy rule weights, membership function positions, and control output ratios. The specific optimization objective function is as follows:

[0048]

[0049] Where λ is the energy consumption weighting coefficient for balancing control accuracy and execution power, and T is the frequency modulation evaluation time window;

[0050] S44: Based on the optimal controller parameters obtained in step S43, the system dynamically adjusts the frequency deviation Δf and the frequency change rate in real time. Generate a power adjustment command u(t) and output it to the corresponding resource to participate in frequency modulation.

[0051] As another improvement of the present invention, the fuzzy logic control process in step S41 is specifically as follows:

[0052]

[0053] Where u(t) is the active power output adjustment value of downstream resources, and FLC(·) is a fuzzy logic control function that combines fuzzy rules with membership functions.

[0054] As a further improvement of the present invention, the frequency control result in step S5 includes, but is not limited to, the calculation of the frequency deviation integral index, wherein the calculation method of the frequency deviation integral index is as follows:

[0055]

[0056] Among them, J f The cumulative frequency deviation index is T, the evaluation time window is T, and Δf(t) = f ref -f(t) represents the system frequency deviation, f ref The reference frequency is set for the system, and f(t) is the real-time frequency measured.

[0057] The resource utilization rate indicators include resource utilization rate, and the calculation methods for various resource utilization rates are as follows:

[0058]

[0059] Among them, U i It is the utilization rate of the i-th type of resource. The adjusted power already used. This indicates the capacity that is currently available but not yet used.

[0060] As a further improvement of the present invention, in step S5, when insufficient control performance or unreasonable resource configuration is detected, the control strategy is corrected, including but not limited to: adjusting the droop coefficient and fuzzy rule weights of the fuzzy controller, updating the resource weight coefficients in the intelligent optimization scheduling model, or optimizing the weights of the frequency modulation objective function.

[0061] Compared with existing technologies, this invention offers the following advantages: First, it provides a frequency stability control method for new energy access to offshore energy platforms. By constructing a unified frequency response modeling framework, it achieves standardized modeling of heterogeneous frequency regulation resources such as energy storage systems, virtual inertial new energy sources, high-voltage direct current, pumped storage, and interruptible loads, thereby improving the system's coordinated response capability under frequency disturbances. Second, it introduces a disturbance identification mechanism based on RoCoF and nadir, combined with a frequency prediction model, to accurately determine the disturbance level, supporting a "fast-medium-slow" hierarchical control strategy and enabling efficient hierarchical resource allocation. Third, it integrates fuzzy logic control and whale optimization algorithms to establish an adaptive parameter tuning method for frequency fluctuation scenarios, enhancing the control system's adaptability to nonlinear disturbances and its global optimization capability. Finally, by constructing a closed-loop feedback mechanism, it evaluates the frequency regulation effect and resource utilization rate in real time, dynamically correcting the control strategy and parameter configuration, achieving self-diagnosis, self-optimization, and continuous stable operation of the frequency control system. This invention significantly improves the frequency control accuracy, response speed, and scheduling intelligence level of offshore energy platforms under complex operating conditions, effectively ensuring system operational safety and economy. Attached Figure Description

[0062] Figure 1 This is a flowchart of the steps of the method of the present invention;

[0063] Figure 2 This is a schematic diagram of the secondary correction and closed-loop optimization of the present invention. Detailed Implementation

[0064] The present invention will be further illustrated below with reference to the accompanying drawings and specific embodiments. It should be understood that the following specific embodiments are for illustrative purposes only and are not intended to limit the scope of the invention.

[0065] Example 1

[0066] A frequency stability control method for new energy access to offshore energy platforms is proposed. This method improves the real-time performance of the system's frequency control through unified modeling of frequency regulation resources, disturbance identification and safety indicator evaluation, hierarchical response control mechanism, adaptive fuzzy logic controller design, and parameter tuning and closed-loop scheduling feedback based on intelligent optimization algorithms. Specifically, as follows... Figure 1 As shown, it includes the following steps:

[0067] Step S1: Construct a unified frequency response modeling framework for multi-source frequency modulation resources.

[0068] To achieve coordinated control of various adjustable frequency resources (energy storage, new energy sources, HVDC, pumped storage, interruptible loads, etc.) on offshore energy platforms, a unified frequency regulation response modeling framework needs to be established. This step identifies resource types, extracts key control parameters, and constructs standardized frequency-power response models. State identification and parameter extraction are performed for frequency regulation resources such as battery energy storage systems (BESS), virtual inertia-controlled wind / photovoltaic systems (VIC-Wind / PV), HVDC, pumped storage (PHS), and interruptible loads, establishing their frequency-power response models. Fast / slow response characteristics are uniformly represented using Laplace frequency domain transfer functions, and a total system frequency response model is constructed by superimposing these functions and encapsulating it into a dictionary structure to support programmatic calls. Specifically, the steps include:

[0069] S11: Extract relevant parameters of resources such as battery energy storage systems (BESS), wind / photovoltaic systems with virtual inertia control (VIC-PV / VIC-Wind), high-voltage direct current regulating equipment (HVDC), and pumped storage systems (PHS) from the real-time monitoring system of the offshore platform. The parameters include, but are not limited to, captured rated capacity, current output, standby regulating capacity, response delay, and whether it can be enabled. It also includes querying the operating status of energy storage such as state of charge, voltage level, and control strategy, and outputting these data into a resource status table in a unified format for subsequent modeling and scheduling.

[0070] S12: Based on the control characteristics of different types of resources, frequency response models for frequency regulation resources are established respectively. Among them, for energy storage systems and fast-regulating devices such as HVDC, a linear function of frequency deviation is used for modeling; for VIC-Wind / PV systems, a joint model of virtual inertia and droop control is established by combining the rate of frequency change; for delayed-response devices such as pumped storage and traditional generators, a first-order hysteresis model in the Laplace domain is used to represent their inertia and delay characteristics.

[0071] To fully characterize the dynamic response behavior of various resources, the model is represented in both the time domain (with t as the variable) and the frequency domain (with s as the variable). In subsequent control strategies, it will be uniformly converted to the frequency domain form for easier unified analysis and scheduling optimization.

[0072] Energy storage systems and high-voltage direct current (HVDC) regulators possess rapid frequency regulation capabilities. Their regulation power response can be modeled as a linear function of the frequency deviation, as follows:

[0073]

[0074] Among them, △P fast (t) The adjusted output power of the energy storage system at time t, k b For frequency response gain coefficient, SoC min and SoCmax These represent the minimum and maximum allowable states of charge of the energy storage system, respectively. △f(t)=f ref -f(t) represents the system frequency deviation, f ref f(t) is the reference frequency set for the system, and f(t) is the real-time frequency measured.

[0075] New energy sources such as wind power and photovoltaics, which possess virtual inertia control capabilities, also have rapid frequency adjustment capabilities, and their response models are represented as follows:

[0076]

[0077] Among them, VP VIC (t) represents the output regulation power of the virtual inertial system, k p and k d These represent the frequency droop control coefficient and the virtual inertia control coefficient, respectively. It is the rate of change of frequency deviation.

[0078] For pumped storage systems and diesel generators, which possess inertial and response delay characteristics, a frequency domain model is more conducive to overall system modeling and analysis. The first-order response model with delay is expressed as follows:

[0079]

[0080] Where s is the complex frequency domain variable under the Laplace transform, and VP i (s) represents the frequency domain output power of the i-th type of frequency modulation resource, Δf(s) represents the frequency deviation signal, and K i T represents the power adjustment caused by a unit change in frequency. i Let τ be the first-order inertial time constant. i To respond to the delay, The transfer function form representing pure delay.

[0081] S13: Utilizing the linear superposition property of the Laplace transform, the frequency domain response functions of various resources are uniformly transformed into frequency-power transfer function forms. Based on parameters such as adjustment gain, inertia time constant, and response delay, a system overall frequency response model is constructed through superposition for subsequent unified scheduling and control analysis.

[0082] The Laplace transform possesses excellent linear superposition properties and advantages in frequency domain analysis, making it suitable for representing dynamic characteristics such as system delay, inertia, and gain. Furthermore, it easily combines the frequency response behaviors of different resources. Therefore, in this step, the response models of various resources are standardized into frequency response transfer function forms, and then superimposed in the frequency domain to construct the overall system response model. The total frequency response power can be expressed as:

[0083]

[0084] Among them, G i (s) represents the frequency form of the response model corresponding to the i-th resource type, including gain, inertia time constant, delay, etc.; ω i The scheduling weight of the i-th resource represents its priority or scheduling proportion in the coordinated frequency regulation.

[0085] S14: The resource parameters and transfer functions generated during the modeling process are stored in a dictionary structure for programmed scheduling control, enabling the frequency controller to quickly retrieve resource states and assess adjustment capabilities. This dictionary supports the controller in quickly finding resource states and calculating adjustment capabilities in real time. It also supports dynamic updates and scalable invocation, providing a modeling foundation and interface support for hierarchical collaborative frequency control strategies. An example dictionary is shown below:

[0086]

[0087]

[0088] Step S2: Design a disturbance sensing and frequency security index calculation mechanism.

[0089] During the operation of offshore energy platforms, frequency security is affected by disturbances such as load fluctuations, new energy output fluctuations, and equipment switching. To achieve rapid response and tiered control, it is necessary to sense the degree of system disturbance in real time and calculate the rate of change of frequency (RoCoF) and the key frequency security indicator of the lowest frequency point (nadir) to provide a basis for subsequent control. Therefore, in this step, by combining frequency deviation, rate of change (RoCoF), and prediction of the lowest frequency point (nadir), the occurrence and level of disturbance are determined in real time. Based on the differential algorithm and an approximate linear model, the frequency change trend is dynamically estimated, providing a precise triggering basis for the control response strategy. Specifically:

[0090] S21: Based on the frequency data collected by the system's main control center, the differential algorithm is used to detect frequency deviation changes in real time, calculate the frequency deviation and rate of change, and determine whether a disturbance event has occurred. A disturbance event is triggered when any of the following conditions are met: 1) The frequency deviation exceeds the static limit (|Δf(t)|>Δf th 2) RoCoF exceeds the threshold 3) The rate of change in new energy power exceeds the fluctuation limit. This is then determined to be a disturbance, triggering emergency frequency control. Where, △f th R is the frequency deviation trigger threshold. th P is the threshold for the rate of change of frequency. VRE The sum of renewable energy power, where α represents the fluctuation threshold.

[0091] S22: Based on a sliding time window and median filtering, noise preprocessing is performed on the sampled frequency data to improve the stability of RoCoF calculation. The rate of change of the system frequency in the early stage of disturbance reflects its inertia level. Median filtering and differential algorithms are used to estimate the rate of frequency change in real time.

[0092]

[0093] Where Δt is the sampling time interval; if the absolute value of RoCoF is greater than R th If the frequency is high, it is determined to be a high-intensity disturbance, and millisecond-level fast response resources are prioritized for frequency suppression.

[0094] S23: To predict in advance the lowest point the system frequency might drop to, an approximate frequency change model is constructed based on the linear frequency decrease assumption, combining the current RoCoF value with the safety threshold Δf. safe Predict the occurrence time and minimum value of nadi r.

[0095] Using a frequency linear variation approximation model, the occurrence time t of nadir is estimated based on the current RoCoF value. nadir With frequency minimum value f nadir :

[0096]

[0097] f nadir =f ref -RoCoF·t nadir

[0098] Among them, △f safe For the safety margin of frequency offset, The time position of the undamped sine wave at the first extreme point is used as a conservative estimation coefficient.

[0099] If the predicted result is lower than the system frequency safety limit, emergency frequency adjustment resources will be triggered immediately to ensure that the system frequency does not fall out of the allowable range.

[0100] S24: Based on the current frequency deviation, rate of change, and nadir prediction results, the disturbance intensity is classified into three levels: mild, moderate, and severe, each corresponding to a different control response strategy. Mild disturbances can be observed and waited for; moderate disturbances will activate some fast response resources; and severe disturbances will enter emergency control mode, fully utilizing all available frequency modulation resources.

[0101] Step S3: Construct a "fast-medium-slow" hierarchical frequency modulation control strategy.

[0102] Based on the disturbance level and the response characteristics of various resources, a hierarchical frequency regulation control strategy of "fast-medium-slow" is constructed. Fast response resources (energy storage, HVDC), medium-speed resources (PHS, diesel engine) and slow resources (interruptible load) are scheduled in stages for coordinated regulation. The frequency status is continuously monitored and the control parameters are dynamically corrected to achieve efficient coordinated use of heterogeneous frequency regulation resources and rapid suppression of frequency deviation.

[0103] S31: After the system detects a sudden frequency drop and determines it to be a medium-to-high intensity disturbance, it immediately mobilizes fast resources with millisecond-level response capabilities, including battery energy storage systems, high-voltage DC regulation equipment, and wind / photovoltaic systems with virtual inertia control capabilities, to participate in frequency regulation. Based on the frequency deviation and frequency response model, it outputs regulation power to achieve initial frequency suppression and rapid stabilization. The regulation method adopts the ΔP proposed in step S1. fast (t) and VP VIC (t) Frequency-power response model.

[0104] S32: After the fast-response resources stabilize, the system enters the first-order frequency regulation stage, at which point medium-speed resources such as pumped storage systems and diesel generators are utilized. These resources typically have slower response times but larger regulation capacities, and are regulated using the delayed first-order response model VP proposed in step S1. i (s), calculate its optimal commissioning time and adjustment range to support the frequency recovery process.

[0105] S33: If the system frequency recovery speed is slow or the early warning frequency regulation capability is insufficient, then slower resources (such as interruptible loads and low-priority loads) are further invoked to participate in frequency regulation control. Based on load importance classification, a linear weighting function is used to determine the power cut-off and load selection, minimizing the impact on system function and production, and achieving flexible load adjustment. The load shedding strategy for interruptible loads can be represented by the following linear weighting model:

[0106]

[0107] Among them, △P cut Represents the total load power currently being cut off, where n is the total number of loads that can be cut off, and c i δ is the unit power of the i-th load. i This is a load shedding status variable (a value of 1 indicates shedding, and 0 indicates retention). Loads are sorted according to their importance or scheduling level, with lower-importance loads being shedding first to minimize disruption to system operation and production.

[0108] S34: After the above resources are introduced in sequence, the control system continuously monitors the frequency change trajectory to determine whether the current frequency has recovered to the permitted operating range.

[0109] f min<f(t)<f max

[0110] Where f(t) is the system frequency at the current moment, f min and f max This represents the operating frequency boundary threshold.

[0111] If the frequency still does not meet the requirements, the adjustment coefficient (k) will be adjusted according to the actual deviation. b k p and k d ), power adjustment range K i And by changing load shedding strategies, secondary corrections and closed-loop optimizations of the control strategy can be achieved, such as... Figure 2 As shown.

[0112] Step S4: Design a fuzzy logic controller to handle nonlinear disturbances.

[0113] By utilizing fuzzy logic control and intelligent optimization algorithms, control parameters are adaptively adjusted based on the real-time frequency status of the system, achieving fine-grained scheduling of frequency modulation resources and globally optimal frequency control performance. The specific implementation is as follows:

[0114] S41: A fuzzy controller is used to handle nonlinear and uncertain problems in frequency control. Its basic structure includes two input variables and one output variable. The inputs are typically the frequency deviation Δf and the rate of change of frequency. The output is the power adjustment command u(t). The two input variables are divided into seven fuzzy language sets: NB (large negative), NM (medium negative), NS (small negative), ZE (zero), PS (small positive), PM (medium positive), and PB (large positive), corresponding to seven triangular membership functions used to characterize the degree of fuzziness of the input state under different disturbance levels. The fuzzy logic control process is as follows:

[0115]

[0116] Where u(t) is the active power output adjustment value of downstream resources, and FLC(·) is a fuzzy logic control function that combines fuzzy rules with membership functions.

[0117] S42: Design a fuzzy inference rule table to map the input language set to the output control action, and realize inference decision-making through a fuzzy rule base. The inference output result is defuzzified using the centroid method to obtain the precise adjustment power value, which is then output to fast response resources (such as BESS, HVDC, etc.) to execute control commands, ensuring the flexibility and continuity of the system response.

[0118] The partitioning of the fuzzy language set in this embodiment is illustrated below:

[0119] Language variables symbol Value range (illustrated) Engineering meaning NB Negative Big [-1.0,-0.75,-0.5] Extremely large frequency deviation NM NegativeMedium [-0.75,-0.5,-0.25] Large frequency deviation NS Negative Small [-0.5,-0.25,0.0] Slight frequency deviation ZE Zero [-0.25,0.0,+0.25] The frequency is basically stable PS Positive Small [0.0,+0.25,+0.5] Slightly high frequency PM Positive Medium [+0.25,+0.5,+0.75] The frequency is significantly higher. PB Positive Big [+0.5,+0.75,+1.0] Severe overclocking

[0120] The specific explanations of the language variables are as follows:

[0121] NB: When the frequency exceeds the normal value and continues to rise, the output power should be reduced rapidly and significantly to suppress the frequency surge.

[0122] NM: When the frequency is significantly too high and the upward trend is strengthening, the output power should be appropriately reduced for adjustment.

[0123] NS: When the frequency is slightly higher and there is an upward trend, the output power can be slightly reduced to control it.

[0124] ZE: The frequency is basically stable with minimal fluctuations and requires no adjustment.

[0125] PS: When the frequency is slightly low and slowly decreasing, the output power should be slightly increased to maintain stability.

[0126] PM: When the frequency is significantly low and continues to decline, the output power should be increased by a moderate amount.

[0127] PB: When the frequency drops significantly and the rate of decline is rapid, the output power should be increased rapidly to intervene.

[0128] Furthermore, the triangular membership function corresponding to each fuzzy language set is defined as follows:

[0129]

[0130] Where, μ A (x) represents the membership degree of variable x in fuzzy set A; a, b, and c represent the left endpoint, peak, and right endpoint positions of the language set, respectively; x is the input variable, which can be Δf or

[0131] Next, the controller outputs the fuzzy control result based on the rule inference table, and uses the centroid method to defuzzify it, obtaining the final accurate output value:

[0132]

[0133] Among them, u * This is the controller output value after defuzzification.

[0134] S43: To further improve control performance, the Whale Optimization Algorithm (WOA) is used to search for the optimal combination of controller parameters, including fuzzy rule weights, membership function positions, and control output ratios. The objective function is to minimize the overall frequency modulation performance index, comprehensively considering frequency deviation, regulation response energy consumption, and system robustness. The optimal controller configuration is obtained through iterative search. The optimization objective function is to minimize the overall frequency modulation performance index:

[0135]

[0136] Where λ is the energy consumption weighting coefficient used to balance control accuracy and execution power, and T is the frequency modulation evaluation time window. The optimization algorithm finds the controller parameter combination that minimizes the objective function J in the solution space through iterative search, thereby achieving the optimal trade-off between frequency control and regulation costs.

[0137] S44: After obtaining the optimal controller parameters, the system adjusts the parameters in real time according to Δf or The system generates a control variable u(t) and outputs it to specific resources (such as BESS, HVDC, or PHS) to participate in frequency regulation. Simultaneously, it feeds back the current frequency status, controller output, and system resource availability to the hierarchical scheduling module, achieving closed-loop linkage between control, scheduling, and feedback, thereby enhancing the adaptability and global coordination capabilities of the control system.

[0138] Step S5: Establish a real-time evaluation and feedback mechanism.

[0139] The system performs real-time evaluation of the frequency control performance to determine whether it meets frequency safety requirements, and provides feedback to correct control parameters or scheduling strategies when necessary, thereby achieving closed-loop optimization and continuous assurance of system stability.

[0140] S51: After the system performs frequency modulation control, it continuously monitors key frequency indicators, including the current frequency f(t), frequency deviation Δf(t), frequency change rate (RoCoF), and the predicted lowest frequency point (nadir). The monitoring objective is to determine whether the system has entered the safe frequency range.

[0141] S52: Compare the current frequency change curve with the preset target response trajectory (such as the expected recovery path or ideal response curve) to evaluate the controller's adjustment effect and analyze whether the controller or scheduling strategy has achieved the expected goal. Define a frequency deviation integral performance index to quantify the cumulative deviation of the frequency modulation response and reflect the stability and timeliness of the system adjustment.

[0142]

[0143] Among them, J f T represents the cumulative frequency deviation index, and T is the evaluation time window.

[0144] S53: Statistics on the actual usage of various frequency regulation resources, calculating resource utilization rate, including the ratio between the requested regulation power and the available capacity. If there is insufficient resource usage or uneven allocation, it will be identified as a state where resource allocation needs optimization.

[0145] Check the actual usage effect of various frequency modulation resources and calculate the utilization rate of each resource. The resource utilization rate can be used to determine whether there is insufficient or uneven allocation of resources, and to assist in subsequent resource reconfiguration.

[0146]

[0147] Among them, U i It is the utilization rate of the i-th type of resource. The adjusted power already used. This indicates the capacity that is currently available but not yet used.

[0148] S54: When insufficient control performance or unreasonable resource allocation is detected, the system will automatically correct control parameters or strategies based on feedback results, including but not limited to: 1. Adjusting the droop coefficient and fuzzy rule weights of the fuzzy controller; 2. Updating the resource weight coefficients in the intelligent optimization scheduling model; 3. Optimizing the weights of the frequency modulation objective function. This process enables the system to have self-evaluation, self-adaptation, and self-correction closed-loop control capabilities, ensuring frequency stability even under complex disturbances or changes in the operating environment.

[0149] In summary, this invention integrates various frequency regulation resources, including energy storage systems, virtual inertia renewable energy sources, high-voltage direct current (HVDC), interruptible loads, and pumped storage, to construct a unified frequency response modeling framework. Through a disturbance sensing mechanism, a hierarchical collaborative control strategy is implemented, incorporating an adaptive fuzzy controller and intelligent optimization algorithms to achieve dynamic coordination between resource regulation capabilities and response delays. Furthermore, by combining the minimum frequency point (nadir) and the rate of change of frequency (RoCoF) constraints, an optimized scheduling model for multiple scenarios is established. This ensures frequency security while reducing frequency regulation costs, improving the frequency control accuracy and rapid response capability of offshore platforms under complex operating conditions, and enhancing the system's intelligence and operational resilience.

[0150] It should be noted that the above content merely illustrates the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. For those skilled in the art, various improvements and modifications can be made without departing from the principle of the present invention, and all such improvements and modifications fall within the scope of protection of the claims of the present invention.

Claims

1. A frequency stability control method for new energy access to offshore energy platforms, characterized in that... It includes at least the following steps: S1: Perform state identification and parameter extraction on frequency regulation resources, and establish their frequency-power response model. The frequency regulation resources include at least battery energy storage systems, wind power / photovoltaic systems with virtual inertia control, high voltage DC equipment, pumped storage systems, and interruptible loads. The frequency domain response functions of various frequency regulation resources are converted into frequency-power transfer function forms, and after superimposing to construct the overall frequency response model of the system, they are encapsulated into a dictionary structure. S2: Based on frequency safety indicators, the occurrence and level of disturbances are determined in real time. Based on differential algorithms and approximate linear models, the frequency change trend is dynamically estimated, and the disturbance level is classified according to the frequency change trend. The frequency safety indicators include at least frequency deviation, rate of change, and prediction of the lowest frequency point. S3: Based on the disturbance level and the response characteristics of various resources, a "fast-medium-slow" hierarchical frequency modulation control strategy is constructed to coordinate the adjustment of different types of resources in stages, monitor the frequency status and dynamically correct the control parameters; S4: Construct a fuzzy logic controller to handle nonlinear disturbances. With frequency deviation and rate of change as input, construct seven types of fuzzy language sets and membership functions to realize frequency modulation power fuzzy control, and introduce the whale optimization algorithm to tune and optimize the parameters of the fuzzy controller. S5: Construct a real-time evaluation and feedback mechanism to dynamically adjust the control strategy based on the frequency control results and resource utilization indicators.

2. The frequency stability control method for new energy access to offshore energy platforms as described in claim 1, characterized in that: Step S1 specifically includes the following steps: S11: Extract the operating parameters of various frequency regulation resources. The frequency regulation resources include at least battery energy storage systems, wind power / photovoltaic systems with virtual inertia control, high voltage DC equipment, pumped storage systems, and interruptible loads. The operating parameters generate a resource status data table in a unified format, including but not limited to rated capacity, real-time output power, standby regulation capacity, response delay, state of charge, voltage level, and control strategy. S12: Construct frequency response models for various resources based on their control characteristics. For battery energy storage systems and high-voltage DC equipment, a linear function of frequency deviation is used for modeling. For wind power / photovoltaic systems with virtual inertia control, a joint model of virtual inertia and droop control is established by combining the rate of frequency change. For pumped storage systems, a first-order hysteresis model in the Laplace domain is used to represent their inertia and delay characteristics. S13: Utilizing the linear superposition property of the Laplace transform, the frequency domain response functions of various resources are uniformly transformed into frequency-power transfer function forms, and the system's overall frequency response model is constructed by superposition; the overall frequency response power is specifically: Among them, G i (s) represents the frequency form of the response model corresponding to the i-th type of resource, ω i Let f(s) be the scheduling weight of the i-th resource, and Δf(s) be the frequency deviation signal. S14: Encapsulate the running parameters and transfer functions of the above resources into a dictionary structure for storage. The dictionary structure supports dynamic updates and scalable calls.

3. The frequency stability control method for new energy access to offshore energy platforms as described in claim 2, characterized in that: In step S12, the battery energy storage system and the high-voltage DC equipment are modeled using a linear function of frequency deviation, specifically as follows: Among them, △P fast (t) The adjusted output power of the energy storage system at time t, k b For frequency response gain coefficient, SoC min and SoC max These represent the minimum and maximum allowable states of charge of the energy storage system, respectively, Δf(t) = f ref -f(t) represents the system frequency deviation, f ref The reference frequency is set for the system, and f(t) is the real-time frequency measured. The wind / photovoltaic system with virtual inertia control establishes a joint model of virtual inertia and droop control based on the frequency change rate, specifically as follows: Among them, VP VIC (t) represents the output regulation power of the virtual inertial system, k p and k d These represent the frequency droop control coefficient and the virtual inertia control coefficient, respectively. It is the rate of change of frequency deviation; The pumped storage system uses a first-order hysteresis model in the Laplace domain to represent its inertia and delay characteristics. The specific first-order delay response model is as follows: Where s is the complex frequency domain variable under the Laplace transform, and VP i (s) represents the frequency domain output power of the i-th type of frequency modulation resource, Δf(s) represents the frequency deviation signal, and K i T represents the power adjustment caused by a unit change in frequency. i Let τ be the first-order inertial time constant. i To respond to the delay, The transfer function form representing pure delay.

4. The frequency stability control method for new energy access to offshore energy platforms as described in claim 1, characterized in that: In step S2, the frequency deviation change is detected in real time using a differential algorithm to determine whether a disturbance event has occurred. Specifically, if the static deviation exceeds the limit |Δf(t)|>Δf th RoCoF exceeded limits Or the rate of change in energy power exceeds the fluctuation threshold. This is determined to be a disturbance, triggering emergency frequency control; where △f th R is the frequency deviation trigger threshold. th P is the threshold for the rate of change of frequency. VRE The sum of renewable energy power, where α represents the fluctuation threshold; The frequency change rate RoCoF is estimated in real time using median filtering and a differential algorithm, specifically: Where Δt is the sampling time interval; if the absolute value of RoCoF is greater than R th If the disturbance is high-intensity, then the fast response resources will be prioritized. Using a frequency linear variation approximation model, the time t for the occurrence of the lowest frequency point is estimated based on the current RoCoF value. nadir With frequency minimum value f nadir : f nadir =f ref -RoCoF·t nadir Among them, △f safe For the safety margin of frequency offset, The time position of the undamped sine wave at the first extreme point is given. If the predicted result is lower than the system frequency safety limit, emergency frequency regulation resources will be triggered immediately.

5. The frequency stability control method for new energy access to offshore energy platforms as described in claim 3, characterized in that: The "fast-medium-slow" layered frequency modulation control strategy in step S3 specifically includes the following steps: S31: When a medium-to-high intensity disturbance occurs, immediately call up rapid resources, including but not limited to battery energy storage systems and high-voltage DC equipment. Based on their frequency-power response model, output regulating power to achieve initial frequency suppression and rapid stabilization. S32: After the rapid resources stabilize, the frequency first-order adjustment stage is entered, and medium-speed resources are called up. The medium-speed resources include, but are not limited to, pumped storage systems. The optimal commissioning time and adjustment range are calculated based on their first-order lag model to support the frequency recovery. S33: If the system frequency recovery speed is slow or the early warning frequency adjustment capability is insufficient, slow resources will be called. The slow resources include, but are not limited to, interruptible loads. They will be sorted according to their importance or scheduling level, and loads with low importance will be cut off first to minimize the interference to system operation and production. S34: The control system continuously monitors the frequency change trajectory and determines whether the current frequency has returned to the permissible operating range. If the frequency still does not meet the requirements, the adjustment coefficient (k) is adjusted according to the actual deviation. b k p and k d ), power adjustment range K i In addition, the load shedding strategy can be changed to achieve secondary correction and closed-loop optimization of the control strategy.

6. The frequency stability control method for new energy access to offshore energy platforms as described in claim 5, characterized in that: In step S33, the load shedding strategy for interruptible loads is represented by the following linear weighted model: Among them, △P cut Represents the total load power currently being cut off, where n is the total number of loads that can be cut off, and c i δ is the unit power of the i-th load. i It is a load shedding state variable.

7. The frequency stability control method for new energy access to offshore energy platforms as described in claim 1, characterized in that: Step S4 specifically includes the following steps: S41: Construct a fuzzy controller whose basic structure includes two input variables and one output variable. The input variables are the frequency deviation Δf and the rate of change of frequency. The output variable is the power adjustment command u(t), and the two input variables are divided into 7 fuzzy language sets: NB (large negative), NM (medium negative), NS (small negative), ZE (zero), PS (small positive), PM (medium positive), PB (large positive), corresponding to 7 triangular membership functions; S42: Design a fuzzy inference rule table to map the input language set to the output control action, and realize inference decision-making through the fuzzy rule base; the inference output result is defuzzified using the centroid method to obtain the precise adjustment power value, and output to the fast resource execution control command. S43: The optimal parameter combination of the controller is searched using the whale optimization algorithm. The optimal parameter combination includes at least fuzzy rule weights, membership function positions, and control output ratios. The specific optimization objective function is as follows: Where λ is the energy consumption weighting coefficient for balancing control accuracy and execution power, and T is the frequency modulation evaluation time window; S44: Based on the optimal controller parameters obtained in step S43, the system dynamically adjusts the frequency deviation Δf and the frequency change rate in real time. Generate a power adjustment command u(t) and output it to the corresponding resource to participate in frequency modulation.

8. The frequency stability control method for new energy access to offshore energy platforms as described in claim 7, characterized in that: The fuzzy logic control process in step S41 is as follows: Where u(t) is the active power output adjustment value of downstream resources, and FLC(·) is a fuzzy logic control function that combines fuzzy rules with membership functions.

9. The frequency stability control method for new energy access to offshore energy platforms as described in claim 1, characterized in that: The frequency control results in step S5 include, but are not limited to, the calculation of the frequency deviation integral index. The specific calculation method for the frequency deviation integral index is as follows: Among them, J f The cumulative frequency deviation index is T, the evaluation time window is T, and Δf(t) = f ref -f(t) represents the system frequency deviation, f ref The reference frequency is set for the system, and f(t) is the real-time frequency measured. The resource utilization rate indicators include resource utilization rate, and the calculation methods for various resource utilization rates are as follows: Among them, U i It is the utilization rate of the i-th type of resource. The adjusted power already used. This indicates the capacity that is currently available but not yet used.

10. The frequency stability control method for new energy access to offshore energy platforms as described in claim 1, characterized in that: In step S5, when insufficient control performance or unreasonable resource configuration is detected, the control strategy is corrected, including but not limited to: adjusting the droop coefficient and fuzzy rule weights of the fuzzy controller, updating the resource weight coefficients in the intelligent optimization scheduling model, or optimizing the weights of the frequency modulation objective function.

Images