Distributed energy storage cooperative control method and system under uncertain source load environment

Abstract

The invention relates to a distributed energy storage cooperative control method and system in a source load uncertain environment. The method comprises the following steps: switching a positive state space model based on non-negative constraint of physical characteristics of an energy storage battery and a dynamic switching behavior of a micro-grid communication network; designing a switching observer based on the switching positive state space model; constructing a self-adaptive power distribution controller; constructing a global closed-loop augmentation system; designing a group of constraint conditions and average residence time conditions based on a linear matrix inequality, and verifying the non-negativity of the closed-loop augmented system matrix; and selecting a common positive Lyapunov function, verifying that the closed-loop augmentation system is globally exponentially stable in a source load uncertain environment, and realizing cooperative control of the distributed energy storage system. The problem of power distribution under communication topology frequent switching and source load random fluctuation is effectively solved, collaborative stable operation of the system on the premise of meeting positive constraint is guaranteed, and the robustness of the system is improved.

Description

Technical Field

[0001] This invention relates to the field of distributed energy storage control engineering technology, and in particular to a distributed energy storage collaborative control method and system under uncertain source-load conditions. Background Technology

[0002] Distributed energy storage systems are a new type of energy regulation method resulting from the combination of smart grid technology and modern power electronics technology. Currently, distributed energy storage technology has been widely applied in microgrids, industrial parks, and island power supply, making a significant contribution to achieving the strategic goals of "carbon peaking and carbon neutrality." The development of distributed energy storage systems has been rapid due to the large-scale integration of intermittent renewable energy sources such as photovoltaics and wind power, and the increasingly higher requirements of power users for power quality and reliability. Simple grid-connected power generation can no longer meet the needs of modern power systems. Distributed energy storage mainly achieves peak shaving and valley filling by storing and releasing surplus power from the source side and transmitting it during peak load periods. However, in the process of dynamic power allocation, it is necessary to strengthen the control of physical constraints and system coordination. This requires establishing a positive switching control procedure covering the entire energy storage cluster to ensure the safety and physical consistency of system operation. On the other hand, in an environment of uncertainty on both the source and load sides, the stable operation of energy storage systems faces a variety of threats and challenges due to the diversity, complexity, and particularity of application scenarios. These threats include not only traditional physical threats such as battery overcharging and over-discharging, and voltage exceeding limits, but also new cybersecurity issues such as increasingly frequent communication topology switching, network attacks, and communication delays. In recent years, cooperative control in complex network environments has become a widely concerned hot topic, leading researchers to propose the theory of positive multi-agent systems. As a system model consistent with physical essence, it strictly defines the non-negativity of state variables, meaning that the battery state of charge, voltage amplitude, and power output at any given time will not exhibit negative values ​​that violate physical meaning. Furthermore, this model defines more stringent constraints and does not rely on nonlinear approximations of system operation. Compared to traditional linear system modeling methods, positive system models, due to their inherent consistency with the non-negativity of physical systems (such as biological populations, chemical reactions, and energy storage batteries), have become a current research hotspot in academia.

[0003] Distributed energy storage systems are characterized by wide distribution, fast response, high cost, and stringent safety requirements. However, in actual operation, the efficiency and stability of power allocation are difficult to achieve due to limitations in battery physical characteristics, harsh communication environments, and random fluctuations in both source and load sides. Because the physical quantities of battery state of charge, voltage, and power in distributed energy storage systems are non-negative and dynamically switching, modeling them as a switching positive multi-agent system is more realistic. Many control methods do not consider the non-negative / positive constraints of the system state when dealing with disturbances during operation (PV output fluctuations, random load changes, communication link failures, etc.), leading to controllers potentially outputting negative voltage or energy commands, posing a serious threat to the physical safety of the energy storage system. Furthermore, the real-world operating environment inevitably encounters random fluctuations in source and load power, causing unknown time-varying net load gaps and reducing the accuracy of power allocation. Simultaneously, the dynamic switching of the communication topology in distributed energy storage systems can also lead to information exchange delays or loss.

[0004] Therefore, the control methods of traditional distributed energy storage systems have problems with the accuracy of power allocation, the physical safety of system operation, and the adaptive capability because they do not consider the non-positive and non-negative constraints of the system state and do not fully take into account the unknown time-varying net load gap caused by random fluctuations in source and load power. Summary of the Invention

[0005] Based on this, in order to solve the above-mentioned technical problems, a distributed energy storage collaborative control method and system under uncertain source-load environment is provided. It can dynamically formulate power allocation strategies that conform to positive constraints, thereby improving the operating efficiency, adaptability and safety level of distributed energy storage system.

[0006] A method for coordinated control of distributed energy storage under uncertain source-load conditions, the method comprising:

[0007] In a microgrid, voltage, current and switching status data of each energy storage unit in a distributed energy storage system are collected. Based on the non-negativity constraint of the physical characteristics of the energy storage battery and the dynamic switching behavior of the microgrid communication network, a switching positive state space model is established for each energy storage unit. The unknown source-load net power gap is used as the time-varying disturbance term of the switching positive state space model.

[0008] A switching observer is designed based on the aforementioned positive state-space model. The switching observer achieves collaborative online estimation of the unmeasurable battery state and the unknown source-load gap through the gain matrix that switches with topology, and outputs state estimates and source-load gap estimates.

[0009] Based on the model parameters of the switching positive state-space model, the state estimate, and the source-load gap estimate, an adaptive power allocation controller is constructed. The adaptive power allocation controller is driven by the state consistency error and adjusts the power and state of each energy storage unit through the cooperative consistency gain matrix, and outputs charging and discharging control commands for each energy storage unit.

[0010] The switching observer, the adaptive power distribution controller, and the switching positive state space model are coupled together to define the global state estimation error and the source-load gap estimation error, and a global closed-loop augmented system containing the dynamics of all energy storage units is constructed using the Kronecker product.

[0011] Design a set of constraints and average residence time conditions based on linear matrix inequalities to verify the non-negativity of the closed-loop augmented system matrix, ensuring that the state of charge and output power of all energy storage units remain non-negative; select common positive Lyapunov functions, combined with the average residence time conditions, to verify that the closed-loop augmented system is globally exponentially stable under uncertain source-load conditions, realizing the collaborative control of the distributed energy storage system.

[0012] In one embodiment, the expression for the switching positive state-space model is:

[0013] , ;

[0014] in, This represents the core state vector of the i-th energy storage unit, which includes the real-time state of charge and the real-time output power. To control the input vector; The observable output vector; The vector representing the net power gap of the unknown source load; The switching signal takes values ​​from a finite set. ; The Metzler matrix, This represents the input matrix that changes with the switching signal. This represents the uncertainty source influence matrix that varies with the switching signal. and These represent the output matrix and feedforward matrix, respectively, which vary with the switching signal, and each matrix is ​​greater than or equal to 0. and They represent 3D vector space.

[0015] In one embodiment, the expression for the switching observer is:

[0016] ,

[0017] ,

[0018] ;

[0019] in, This represents the state estimate of the i-th energy storage unit; This represents the net power gap of unknown sources and loads in a microgrid. The estimated value; This represents the observer's output estimate; This represents the state observer gain matrix; This represents the auxiliary system matrix that is dynamically related to the unknown source load; and This represents the adaptive gain matrix with unknown parameters that change with topology switching.

[0020] In one embodiment, the expression for the adaptive power distribution controller is:

[0021] ;

[0022] in, Indicates the current switching signal Below, the adjacency matrix weights corresponding to the communication topology, when energy storage units i and j have a communication link, ; This represents the state consistency error between the i-th energy storage unit and its adjacent j-th energy storage unit; Represents the consensus gain matrix; This represents the state feedback gain matrix.

[0023] In one embodiment, the state estimation error The source-charge gap estimation error represents the deviation between the estimated state of charge or power of the i-th battery module and the actual value. This indicates the prediction bias of the switching observer for the unknown net load gap of the microgrid;

[0024] The dynamic evolution equation of the state estimation error is:

[0025] ;

[0026] The dynamic evolution equation for the source load gap estimation error is:

[0027] .

[0028] In one embodiment, a global closed-loop augmented system incorporating the dynamics of all energy storage units is constructed using the Kronecker product, including:

[0029] Define augmented closed-loop state vector Global state vector State estimation error vector Source load gap estimation error vector ;

[0030] Assume switching signal Using the Kronecker product, the global closed-loop system is obtained as follows:

[0031] ;

[0032] Define global augmenting state The globally closed-loop augmented system is represented as:

[0033] ;

[0034] in, Represents the Kronecker product symbol. This represents the Laplace matrix.

[0035] In one embodiment, the distributed energy storage system is positive and stable when the constraint condition based on the linear matrix inequality is satisfied; the average residence time condition is: ,in, >1 indicates switching gain. >0 represents the attenuation coefficient, and the gain matrix of the switching observer and controller is obtained by solving the constraint conditions based on the linear matrix inequality.

[0036] In one embodiment, the non-negativity of the closed-loop augmented system matrix is ​​verified by proving that the matrix of the closed-loop augmented system is ≥0, and by deriving the error dynamic system through the consistency error transformation, combined with the matrix non-negativity condition, to ensure... ≥0 holds true for all t≥0.

[0037] In one embodiment, the expression for the copositive Lyapunov function is: ,in, , , are all column vectors composed of positive vectors.

[0038] A distributed energy storage collaborative control system for environments with uncertain source and load, the system comprising:

[0039] The switching positive state space model establishment module is used to collect voltage, current and switching state data of each energy storage unit in the distributed energy storage system in the microgrid, and establish a switching positive state space model for each energy storage unit based on the non-negativity constraint of the physical characteristics of the energy storage battery and the dynamic switching behavior of the microgrid communication network. The unknown source-load net power gap is used as the time-varying disturbance term of the switching positive state space model.

[0040] A switching observer design module is used to design a switching observer based on the switching positive state space model. The switching observer achieves cooperative online estimation of the unmeasurable battery state and the unknown source load gap through the gain matrix that switches with topology, and outputs the state estimate and the source load gap estimate.

[0041] An adaptive power allocation controller construction module is used to construct an adaptive power allocation controller based on the model parameters of the switching positive state space model, the state estimate, and the source-load gap estimate. The adaptive power allocation controller is driven by the state consistency error and adjusts the power and state of each energy storage unit through the cooperative consistency gain matrix, and outputs charging and discharging control commands for each energy storage unit.

[0042] A global closed-loop augmented system construction module is used to couple the switching observer, the adaptive power distribution controller and the switching positive state space model, define the global state estimation error and source-load gap estimation error, and construct a global closed-loop augmented system containing the dynamics of all energy storage units using the Kronecker product.

[0043] The condition design and verification module is used to design a set of constraints and average residence time conditions based on linear matrix inequalities to verify the non-negativity of the closed-loop augmented system matrix, ensuring that the state of charge and output power of all energy storage units remain non-negative. By selecting a common positive Lyapunov function and combining it with the average residence time condition, it is verified that the closed-loop augmented system is globally exponentially stable under uncertain source-load conditions, realizing the coordinated control of the distributed energy storage system.

[0044] The aforementioned distributed energy storage collaborative control method and system under uncertain source-load environments ensures physical consistency in modeling by establishing a positive switching state-space model for each energy storage unit that conforms to the non-negative physical constraints of the battery. By designing a switching observer, a collaborative online estimation of the unmeasurable battery state and unknown source-load gap is achieved through a gain matrix that changes with topology switching. This outputs accurate state and source-load gap estimates, addressing the critical issue of unmeasurable information in practical applications and providing a reliable benchmark for power allocation. The adaptive power allocation controller adjusts the power balancing speed of each unit through a collaborative consistency gain matrix and ensures the stability of individual units through its own state feedback gain matrix. It outputs charging and discharging control commands for each energy storage unit, achieving dynamic power balancing in actual operation. A global closed-loop augmented system is constructed to provide quantitative evidence for subsequent closed-loop analysis and stability verification, facilitating the overall verification of system positivity and stability. Constraints are designed using a common positive Lyapunov function for system verification, ensuring reliable operation during topology switching and source-load fluctuations in practical applications. This improves the operating efficiency, adaptability, and safety level of the distributed energy storage system. Attached Figure Description

[0045] Figure 1 This is a schematic diagram of a distributed energy storage system under uncertain source-load conditions in one embodiment;

[0046] Figure 2 This is a flowchart illustrating a distributed energy storage collaborative control method under uncertain source-load conditions in one embodiment.

[0047] Figure 3 This is a block diagram of a distributed power dynamic control structure based on a switching observer in one embodiment;

[0048] Figure 4 This is a block diagram of a distributed energy storage collaborative control system under uncertain source-load conditions in one embodiment. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0050] The distributed energy storage collaborative control method under uncertain source-load environments provided in this application can be applied to, for example... Figure 1 In the two-layer architecture of the distributed energy storage system shown, namely the physical layer and the information layer, as follows: Figure 1 As shown, the hardware components, external environmental inputs, and communication topology characteristics of the distributed energy storage system intuitively reflect its physical implementation. The physical layer comprises multiple distributed energy storage units (Batteries), each equipped with an independent DC / DC converter and local control unit (LCU), achieving centralized transmission and distribution of electrical energy via a DC bus. For example... Figure 1 As shown, random photovoltaic output and random load demand are incorporated, which together constitute the source-load uncertainty environment of the system; it can realize the storage, release and physical energy transmission of electrical energy, and is the core carrier of power exchange in distributed energy storage systems.

[0051] like Figure 1 As shown, the information layer labels the communication link switching, reflecting the dynamic changes in the communication topology. The switching signal provides a basis for the distributed energy storage system model to adapt to topology changes. Through communication link identifiers such as C1, C2, and C3, the information interaction channels between energy storage units are displayed, providing information transmission support for distributed collaborative control. It can realize the interaction of status information, topology status identification, and control command transmission between energy storage units, and is the core carrier of system collaborative decision-making.

[0052] Sensor data from the physical layer is transmitted upwards to the information layer, providing basic data for control decisions; control commands from the information layer are transmitted downwards to the LCU and DC / DC converter in the physical layer, driving the energy storage unit to perform charging and discharging operations, thus realizing a coordinated closed loop of energy flow in the physical layer and decision support in the information layer.

[0053] In one embodiment, such as Figure 2 As shown, a distributed energy storage collaborative control method under uncertain source-load environment is provided, including the following steps:

[0054] Step 202: Collect voltage, current and switching status data of each energy storage unit in the distributed energy storage system in the microgrid. Based on the non-negativity constraint of the physical characteristics of the energy storage battery and the dynamic switching behavior of the microgrid communication network, establish a switching positive state space model for each energy storage unit and use the unknown source-load net power gap as the time-varying disturbance term of the switching positive state space model.

[0055] For a microgrid system containing N distributed energy storage units, voltage, current, and switching status data are collected. Considering that the state of charge and output power of the energy storage batteries must be non-negative, and that topology switching may occur due to electromagnetic interference or plug-and-play operation of the communication network, a dynamic model of the i-th energy storage node is established.

[0056] In one embodiment, the expression for switching the positive state-space model is:

[0057] , ;

[0058] in, This represents the core state vector of the i-th energy storage unit, which includes the battery's real-time state of charge and real-time output power. This is the control input vector for the i-th energy storage unit, i.e., the charge / discharge regulation command issued by the battery management system; is the observable output vector of the i-th energy storage unit, i.e., the voltage or power data obtained directly by the sensor; The unknown source-load net power gap vector in the microgrid is determined by the intermittent fluctuations of photovoltaic output on the source side and the random switching of user loads on the load side. It is an unknown time-varying target that the system needs to track and balance. The switching signal takes values ​​from a finite set. This indicates that the switching signal corresponds to the current communication topology mode or operating mode of the microgrid; The system state matrix represents the changes in response to the switching signal, describing the self-discharge characteristics and charge / discharge dynamics within the energy storage battery. It is a Metzler matrix, meaning that all off-diagonal elements are non-negative; This represents the input matrix that varies with the switching signal, describing the gain effect of the control command on the battery state. ≥0 means that all elements are non-negative; This represents the uncertain source influence matrix that varies with the switching signal, describing the unknown source-load gap. The weight of the direct influence on the state of the i-th energy storage unit, and ≥0; and Let represent the output matrix and feedforward matrix, respectively, which vary with the switching signal. These matrices describe the mapping relationship between sensor measurements, system state, and unknown source-load gaps, and satisfy the following conditions: ≥0、 ≥0; J represents the number of the distributed energy storage unit, and J represents the set of positive integers. and They represent 3D vector space.

[0059] Step 204: Design a switching observer based on the switching positive state-space model. The switching observer achieves the cooperative online estimation of the unmeasurable battery state and the unknown source-load gap through the gain matrix that switches with the topology, and outputs the state estimate and the source-load gap estimate.

[0060] In one embodiment, due to the net power gap between the source and load Since it cannot be directly measured, and the internal state of some energy storage units is unmeasurable due to sensor limitations, the expression for switching observers is:

[0061] ,

[0062] ,

[0063] ;

[0064] in, This represents the state estimate of the i-th energy storage unit; This represents the net power gap of unknown sources and loads in a microgrid. The estimated value, through adaptive adjustment of the observer, can approximate the real uncertain source load difference in real time, thereby providing a benchmark for power allocation; This represents the observer's output estimate; This represents the state observer gain matrix that changes with topology and is used to correct state estimation errors. This represents the auxiliary system matrix that is dynamically related to the unknown source load; and This represents the adaptive gain matrix of unknown parameters that changes with topology switching, used to dynamically adjust the estimation of source load gaps using the output residuals.

[0065] Step 206: Based on the model parameters of the switching positive state-space model, the state estimate, and the source-load gap estimate, an adaptive power allocation controller is constructed. The adaptive power allocation controller is driven by the state consistency error and adjusts the power and state of each energy storage unit through the cooperative consistency gain matrix, and outputs charging and discharging control commands for each energy storage unit.

[0066] In one embodiment, the expression for the adaptive power distribution controller is:

[0067] ;

[0068] in, Indicates the current switching signal Below, the adjacency matrix weights corresponding to the communication topology, when energy storage units i and j have a communication link, Otherwise, it is 0, indicating the real-time connectivity status of the microgrid communication network; This represents the state consistency error between the i-th energy storage unit and its adjacent j-th energy storage unit, used to drive all energy storage units to tend towards equilibrium; This represents the coordination and consistency gain matrix, which is used to adjust the balancing speed of power distribution among energy storage units to ensure that the state of charge tends to be consistent. This represents the state feedback gain matrix, used to ensure the stability of a single energy storage unit when tracking an unknown source load gap and to prevent state divergence.

[0069] Step 208: Couple the switching observer, adaptive power distribution controller and switching positive state-space model, define the global state estimation error and source-load gap estimation error, and use the Kronecker product to construct a global closed-loop augmented system that includes the dynamics of all energy storage units.

[0070] In one embodiment, the state estimation error This represents the deviation between the estimated state of charge or power of the i-th battery module and the actual value; source-load gap estimation error. This indicates the prediction bias of the switching observer for the unknown net load gap in the microgrid;

[0071] Combining the switching positive state-space model and the expression for the switching observer, the dynamic evolution equation of the state estimation error can be derived as follows:

[0072] ;

[0073] Combining the adaptive rate in the observer switching expression, the dynamic evolution equation of the source load gap estimation error can be derived as follows:

[0074] .

[0075] To analyze the stability and cooperative performance of the entire microgrid system, the Kronecker product is used to construct the augmented state vector of the entire system.

[0076] In one embodiment, a distributed energy storage collaborative control method under uncertain source-load conditions may further include a process of establishing a global closed-loop augmented system, specifically including: defining the augmented closed-loop state vector. Construct the augmented system of the i-th energy storage system, expressed as:

[0077]

[0078] Define the global state vector State estimation error vector Source load gap estimation error vector ; Assume switching signal Using the Kronecker product, the global closed-loop system is obtained as follows:

[0079] ;

[0080] Define global augmenting state Then the global closed-loop augmented system can be represented in a compact form:

[0081] ;

[0082] in, Represents the Kronecker product symbol. This represents the Laplace matrix.

[0083] Step 210: Design a set of constraints and average residence time conditions based on linear matrix inequalities to verify the non-negativity of the closed-loop augmented system matrix, ensuring that the state of charge and output power of all energy storage units remain non-negative; select common positive Lyapunov functions, combined with the average residence time condition, to verify that the closed-loop augmented system is globally exponentially stable under uncertain source-load conditions, thus realizing the coordinated control of the distributed energy storage system.

[0084] In one embodiment, the distributed energy storage system is positive and stable when the constraint condition based on the linear matrix inequality is satisfied; the mean residence time condition is: ,in, >1 indicates switching gain. >0 represents the attenuation coefficient, and the gain matrices of the switching observer and controller are obtained by solving the constraint conditions based on linear matrix inequalities.

[0085] Specifically, design constants and , vector , This makes the following inequality:

[0086]

[0087]

[0088]

[0089]

[0090]

[0091] Design for any If this condition holds true, then the distributed energy storage system is positive and stable under the switching observer and adaptive power distribution controller. The mean residence time condition is... The gain matrices of the designed observer and controller are as follows:

[0092]

[0093] in, Indicates that all elements are 1 3D column vector, Indicates the first One element is 1, and the rest are 0. 3D column vector. Superscript symbol This indicates transpose.

[0094] In one embodiment, the non-negativity of the closed-loop augmented system matrix is ​​verified by proving that the matrix of the closed-loop augmented system is ≥0, and by deriving the error dynamic system through the consistency error transformation, combined with the matrix non-negativity condition to ensure... ≥0 holds true for all t≥0.

[0095] Specifically, the nonnegativity verification process of the system augmented matrix is ​​as follows: Combining the constraints based on linear matrix inequalities, we can obtain... Depend on and It can be seen that it is easy to obtain .Depend on It can be seen that it is easy to obtain .according to achievable Therefore, the closed-loop augmented system is positive.

[0096] The positive analysis process based on consistency error transformation is as follows: to rigorously prove the state of each energy storage unit The non-negativity of the system itself requires further analysis of the system's consistency error dynamics. Define the consistency error vector for the i-th energy storage unit: in, These are the weighting coefficients. Define the global consistency error vector:

[0097]

[0098] Using the Kronecker product, these error vectors can be compactly represented as:

[0099]

[0100] in, It is the Laplace matrix The left eigenvector corresponding to the zero eigenvalue, It is a constant. The key property of this variable is: This transformation converts the absolute state of each energy storage unit into a deviation from the weighted average state. Analyzing the convergence of these deviations is equivalent to analyzing whether the original system state tends to be consistent.

[0101] The derivation of the dynamic equation for the error vector is as follows: Differentiate the above error vector and substitute it into the original system's dynamic equation: After expansion, we get the matrix form: Similarly, there are The expression. Further substitution... and The specific form, after algebraic operations and using properties Thus, a compact error dynamic system can be obtained:

[0102]

[0103] definition Then the above equation can be written as: ,in, .

[0104] Passing conditions , can be obtained ,in, This represents the i-th diagonal element of the Laplace matrix; Represents the element in the j-th row and i-th column of the Laplace matrix ( Based on the proof of the nonnegativity verification process of the system augmented matrix, it has been proven that... and .also, and Therefore, the positivity of distributed energy storage systems is guaranteed.

[0105] In one embodiment, the expression for the copositive Lyapunov function is: ,in, , , are all column vectors composed of positive vectors.

[0106] Specifically, the stability verification process of distributed energy storage systems under uncertain source-load environments includes: selecting a common positive Lyapunov function; calculating the derivative of the Lyapunov function and performing boundedness analysis. Furthermore, substituting the error vector into the dynamic equation derivation process, we can obtain:

[0107] Considering the net power gap between source and load Bounded (determined by the characteristics of the physical system), and has constants. , so that: .

[0108] Passing conditions From the designed gain matrix, we can obtain: Further obtain .in, This represents the system attenuation coefficient, the magnitude of which depends on the design of the controller and observer gains. In each fixed topology mode... Below, the system energy function It converges to a bounded region at an exponential rate.

[0109] Considering the switching behavior of the communication topology, the switching signal The average length of stay meets the following requirements: ,in Take any time point ,right By integrating and using the average dwell time condition, we can obtain the following through recursion: Further consider the previous interval During the switching time The energy may jump, satisfying ,get:

[0110]

[0111] This recurrence relation from the initial time It continues to apply up to the current moment. And using the definition of average residence time, i.e. ,available:

[0112] ,in, For the number of switching times, This is the upper bound constant for the disturbance term. Therefore, when the average residence time satisfies... Sometimes, .when At that time, the final convergence limit of the system's energy function is: Due to the Lyapunov function It eventually converges to a bounded constant, and the system state... and Proportional, therefore the error state It also converges to a bounded region. Specifically, when the average dwell time satisfies... At that time, the system is exponentially asymptotically stable.

[0113] In one embodiment, the block diagram of the distributed power dynamic control structure based on the switching observer is as follows: Figure 3 As shown, external disturbances correspond to Figure 1 The uncertain environment of source and load (PV output fluctuations, random load switching) is input into the system in the form of an unknown source and load net power gap; the switching signal is output by the switching signal generator, which is generated based on the communication topology identification result and adapts to the dynamic changes of the information layer communication topology; the system state variables include core physical quantities such as the state of charge (SOC) and output power of the energy storage unit, providing state feedback for control decisions.

[0114] The switching observer receives system output, switching signals, and control inputs. Its core function is to achieve state estimation and disturbance estimation, and output accurate estimation results to solve the problem of unmeasurable information in practical applications. The adaptive controller establishes a consensus protocol, receives the state and disturbance estimates and switching signals output by the switching observer, and generates targeted control input commands with power balance as the goal. Its core function is to adjust the power distribution of each energy storage unit to ensure coordinated balance and stability of individual units.

[0115] The control input commands generated by the adaptive controller are transmitted to the distributed energy storage system, driving each unit to perform charge and discharge regulation; the system output state variables (SOC, power) are fed back to the switching observer, forming a closed-loop logic of observation-control-feedback, ensuring control accuracy and system stability.

[0116] It should be understood that although the steps in the flowchart above are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowchart above may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0117] In one embodiment, such as Figure 4 As shown, a distributed energy storage collaborative control system under uncertain source-load environment is provided, including: a switching positive state-space model establishment module 410, a switching observer design module 420, an adaptive power distribution controller construction module 430, a global closed-loop augmented system construction module 440, and a condition design and verification module 450, wherein:

[0118] The switching positive state space model establishment module 410 is used to collect voltage, current and switching state data of each energy storage unit in the distributed energy storage system in the microgrid, and establish a switching positive state space model for each energy storage unit based on the non-negativity constraint of the physical characteristics of the energy storage battery and the dynamic switching behavior of the microgrid communication network, and use the unknown source-load net power gap as the time-varying disturbance term of the switching positive state space model.

[0119] The switching observer design module 420 is used to design a switching observer based on the switching positive state space model. The switching observer achieves the cooperative online estimation of the unmeasurable battery state and the unknown source load gap through the gain matrix that switches with the topology, and outputs the state estimate and the source load gap estimate.

[0120] The adaptive power distribution controller construction module 430 is used to construct an adaptive power distribution controller based on the model parameters of the switching positive state space model, the state estimate, and the source-load gap estimate. The adaptive power distribution controller is driven by the state consistency error and adjusts the power and state of each energy storage unit through the cooperative consistency gain matrix, and outputs the charging and discharging control commands of each energy storage unit.

[0121] The global closed-loop augmented system construction module 440 is used to couple the switching observer, the adaptive power distribution controller and the switching positive state space model, define the global state estimation error and the source-load gap estimation error, and construct a global closed-loop augmented system containing the dynamics of all energy storage units using the Kronecker product.

[0122] The condition design and verification module 450 is used to design a set of constraints and average residence time conditions based on linear matrix inequalities to verify the non-negativity of the closed-loop augmented system matrix and ensure that the state of charge and output power of all energy storage units remain non-negative. By selecting a common positive Lyapunov function and combining it with the average residence time condition, it is verified that the closed-loop augmented system is globally exponentially stable under uncertain source-load conditions, thus realizing the coordinated control of the distributed energy storage system.

[0123] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0124] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A distributed energy storage collaborative control method under uncertain source-load conditions, applied to a distributed energy storage system, characterized in that, The method includes: In a microgrid, voltage, current and switching status data of each energy storage unit in a distributed energy storage system are collected. Based on the non-negativity constraint of the physical characteristics of the energy storage battery and the dynamic switching behavior of the microgrid communication network, a switching positive state space model is established for each energy storage unit. The unknown source-load net power gap is used as the time-varying disturbance term of the switching positive state space model. A switching observer is designed based on the aforementioned positive state-space model. The switching observer achieves collaborative online estimation of the unmeasurable battery state and the unknown source-load gap through the gain matrix that switches with topology, and outputs state estimates and source-load gap estimates. Based on the model parameters of the switching positive state-space model, the state estimate, and the source-load gap estimate, an adaptive power allocation controller is constructed. The adaptive power allocation controller is driven by the state consistency error and adjusts the power and state of each energy storage unit through the cooperative consistency gain matrix, and outputs charging and discharging control commands for each energy storage unit. The switching observer, the adaptive power distribution controller, and the switching positive state space model are coupled together to define the global state estimation error and the source-load gap estimation error, and a global closed-loop augmented system containing the dynamics of all energy storage units is constructed using the Kronecker product. Design a set of constraints and average residence time conditions based on linear matrix inequalities to verify the non-negativity of the closed-loop augmented system matrix, ensuring that the state of charge and output power of all energy storage units remain non-negative; select common positive Lyapunov functions, combined with the average residence time conditions, to verify that the closed-loop augmented system is globally exponentially stable under uncertain source-load conditions, realizing the collaborative control of the distributed energy storage system.

2. The distributed energy storage collaborative control method under uncertain source-load environment according to claim 1, characterized in that, The expression for the switching positive state-space model is: ， ； in, This represents the core state vector of the i-th energy storage unit, which includes the real-time state of charge and the real-time output power. To control the input vector; The observable output vector; The vector representing the net power gap of the unknown source load; The switching signal takes values ​​from a finite set. ; The Metzler matrix, This represents the input matrix that changes with the switching signal. This represents the uncertainty source influence matrix that varies with the switching signal. and These represent the output matrix and feedforward matrix, respectively, which vary with the switching signal, and each matrix is ​​greater than or equal to 0. and They represent 3D vector space.

3. The distributed energy storage collaborative control method under uncertain source-load environment according to claim 2, characterized in that, The expression for the switching observer is: ， ， ； in, This represents the state estimate of the i-th energy storage unit; This represents the net power gap of unknown sources and loads in a microgrid. The estimated value; This represents the observer's output estimate; This represents the state observer gain matrix; This represents the auxiliary system matrix that is dynamically related to the unknown source load; and This represents the adaptive gain matrix with unknown parameters that change with topology switching.

4. The distributed energy storage collaborative control method under uncertain source-load environment according to claim 3, characterized in that, The expression for the adaptive power distribution controller is: ； in, Indicates the current switching signal Below, the adjacency matrix weights corresponding to the communication topology, when energy storage units i and j have a communication link, ; This represents the state consistency error between the i-th energy storage unit and its adjacent j-th energy storage unit; Represents the consensus gain matrix; This represents the state feedback gain matrix.

5. The distributed energy storage collaborative control method under uncertain source-load environment according to claim 4, characterized in that, The state estimation error The source-charge gap estimation error represents the deviation between the estimated state of charge or power of the i-th battery module and the actual value. This indicates the prediction bias of the switching observer for the unknown net load gap of the microgrid; The dynamic evolution equation of the state estimation error is: ； The dynamic evolution equation for the source load gap estimation error is: 。 6. The distributed energy storage collaborative control method under uncertain source-load environment according to claim 5, characterized in that, Constructing a global closed-loop augmented system that incorporates the dynamics of all energy storage units using the Kronecker product, including: Define augmented closed-loop state vector Global state vector State estimation error vector Source load gap estimation error vector ; Assume switching signal Using the Kronecker product, the global closed-loop system is obtained as follows: ； Define global augmenting state The globally closed-loop augmented system is represented as: ； in, Represents the Kronecker product symbol. This represents the Laplace matrix.

7. The distributed energy storage collaborative control method under uncertain source-load environment according to claim 6, characterized in that, The distributed energy storage system is positive and stable when the constraint condition based on the linear matrix inequality is satisfied; the average residence time condition is: ,in, >1 indicates switching gain. >0 represents the attenuation coefficient, and the gain matrix of the switching observer and controller is obtained by solving the constraint conditions based on the linear matrix inequality.

8. The distributed energy storage collaborative control method under uncertain source-load environment according to claim 7, characterized in that, The non-negativity of the closed-loop augmented system matrix is ​​verified by proving that the matrix of the closed-loop augmented system is ≥0, and by deriving the error dynamic system through the consistency error transformation, combined with the matrix non-negativity condition, ensuring... ≥0 holds true for all t≥0.

9. The distributed energy storage collaborative control method under uncertain source-load environment according to claim 8, characterized in that, The expression for the common positive Lyapunov function is: ,in, , , are all column vectors composed of positive vectors.

10. A distributed energy storage collaborative control system for uncertain source-load environments, characterized in that, The system includes: The switching positive state space model establishment module is used to collect voltage, current and switching state data of each energy storage unit in the distributed energy storage system in the microgrid, and establish a switching positive state space model for each energy storage unit based on the non-negativity constraint of the physical characteristics of the energy storage battery and the dynamic switching behavior of the microgrid communication network. The unknown source-load net power gap is used as the time-varying disturbance term of the switching positive state space model. A switching observer design module is used to design a switching observer based on the switching positive state space model. The switching observer achieves cooperative online estimation of the unmeasurable battery state and the unknown source load gap through the gain matrix that switches with topology, and outputs the state estimate and the source load gap estimate. An adaptive power allocation controller construction module is used to construct an adaptive power allocation controller based on the model parameters of the switching positive state space model, the state estimate, and the source-load gap estimate. The adaptive power allocation controller is driven by the state consistency error and adjusts the power and state of each energy storage unit through the cooperative consistency gain matrix, and outputs charging and discharging control commands for each energy storage unit. A global closed-loop augmented system construction module is used to couple the switching observer, the adaptive power distribution controller and the switching positive state space model, define the global state estimation error and source-load gap estimation error, and construct a global closed-loop augmented system containing the dynamics of all energy storage units using the Kronecker product. The condition design and verification module is used to design a set of constraints and average residence time conditions based on linear matrix inequalities to verify the non-negativity of the closed-loop augmented system matrix, ensuring that the state of charge and output power of all energy storage units remain non-negative. By selecting a common positive Lyapunov function and combining it with the average residence time condition, it is verified that the closed-loop augmented system is globally exponentially stable under uncertain source-load conditions, realizing the coordinated control of the distributed energy storage system.

Images