Model prediction control method considering pumped storage and battery energy storage collaborative frequency modulation and related device

Abstract

The invention relates to a model prediction control method considering pumped storage and battery energy storage collaborative frequency modulation. The method comprises the following steps: constructing a dynamic model in which a hybrid pumped storage-battery energy storage system participates in secondary frequency modulation; according to the dynamic model of the hybrid pumped storage-battery energy storage system participating in secondary frequency modulation, establishing a state-space equation model equivalent to the dynamic model; discretizing the state-space equation model to obtain a discrete prediction model; according to the discrete prediction model, a target optimization function with the minimum quadratic sum of system frequency deviation is established, and constraint conditions are determined; solving the target optimization function at each current moment to obtain an optimal decision sequence of future N steps; and taking the first step of the optimal decision sequence as an optimal action control instruction of the hybrid pumped storage-battery energy storage system at the current moment, issuing the optimal action control instruction to the hybrid pumped storage-battery energy storage system to participate in frequency modulation, and adjusting the frequency deviation to zero in a relatively short time under the condition of wind-solar power fluctuation.

Description

Technical Field

[0001] The present invention belongs to the field of power system operation and automatic control, and specifically relates to a model predictive control method for frequency regulation taking into account the combined participation of pumped storage and battery energy storage. Background Art

[0002] The inherent fluctuation characteristics of renewable energy generation will lead to power imbalances between sources and loads in the power system on multiple time scales, which poses a severe challenge to maintaining power balance for traditional thermal power units. In order to ensure high power quality, generators need to start and stop frequently to compensate for short-term power changes and thus adjust the frequency. However, high-frequency start and stop operations will accelerate the wear of generators and bring additional operating costs. At the same time, in the future, when thermal power units are gradually replaced by renewable energy, the flexibility regulation resources in the power system are facing a serious shortage. Therefore, it is necessary to rely on more other flexibility resources to ensure the frequency stability of the power system.

[0003] Battery energy storage has become a flexible regulation resource with a very wide range of applications due to its fast ramp response capability, which can provide flexibility and reduce frequency deviation. In addition, in some mountainous areas with abundant water resources, pumped storage has attracted much attention due to its long service life and large capacity. The inventors found that the existing frequency controller designs are more focused on thermal power units or single energy storage as the control object, and have failed to fully explore the flexibility regulation capabilities of hybrid pumped storage-battery energy storage. Therefore, how to design a control strategy so that pumped storage and battery energy storage can collaboratively participate in power system frequency control has become an urgent problem to be solved. Summary of the invention

[0004] Based on this, the present invention provides a model predictive control method that takes into account the coordinated frequency modulation of pumped storage and battery energy storage on the basis of fully tapping the potential of pumped storage and battery energy storage.

[0005] The present invention aims to solve the problem of hybrid pumped storage and new energy storage jointly participating in frequency control to reduce the impact of wind and solar active power fluctuations on the frequency of the power system.

[0006] The technical solution adopted by the present invention is: a model predictive control method taking into account the coordinated frequency modulation of pumped storage and battery energy storage, comprising:

[0007] Construct a dynamic model of a hybrid pumped storage-battery energy storage system participating in secondary frequency regulation;

[0008] According to the dynamic model of hybrid pumped storage-battery energy storage system participating in secondary frequency regulation, an equivalent state space equation model is established;

[0009] Discretize the state space equation model to obtain a discrete prediction model;

[0010] According to the discrete prediction model, the objective optimization function of minimizing the sum of squares of system frequency deviations is established, and the constraint conditions are determined;

[0011] At each current moment, the optimization model composed of the objective optimization function (13) and the constraint condition (14) is solved to obtain the optimal decision sequence; the first step of the optimal decision sequence is taken as the optimal action control instruction of the hybrid pumped storage-battery energy storage system at the current moment, and is sent to the hybrid pumped storage-battery energy storage system to participate in frequency regulation.

[0012] The dynamic model of the hybrid pumped storage-battery energy storage system participating in secondary frequency regulation includes: a transfer function between turbine valve opening and mechanical power, a transfer function of the turbine speed governor, and a battery energy storage transfer function; wherein,

[0013] The transfer function between turbine valve opening and mechanical power is:

[0014]

[0015] Among them, ΔP m represents the output power deviation of pumped storage, ΔG is regarded as the gate opening deviation, s is the Laplace operator, T w is the water starting time constant, which is in the range between 0.5 s and 4.0 s at full load;

[0016] The transfer function of the turbine speed governor is:

[0017]

[0018] Where Δω represents the angular velocity deviation, T gPS is the governor constant of pumped storage in power generation mode, T R is the reset time, R T is the transient droop parameter, R PS is the static droop parameter;

[0019] The battery energy storage transfer function is:

[0020]

[0021] ΔP nL (s)-ΔP m (s)-ΔP B (s) = Δf(s)(D + sM) (4)

[0022] ΔP nL (s) = ΔP Load (s)-ΔP PV (s)-ΔP wind (s) (5)

[0023] Equations (4) and (5) represent the mathematical relationship between frequency deviation and unbalanced power. In the above equations, ΔP B is the output power deviation of the battery storage, u B Indicates automatic power generation control signal; K B and T B Denote the charge / discharge coefficient and time constant respectively; Δf denotes the frequency deviation of the power system; D and M denote the load damping constant and inertia constant respectively; ΔP nL is the net power change; ΔP Load , ΔP PV and ΔP wind They represent the changes in load demand, photovoltaic output power and wind power generation power respectively.

[0024] The state space equation model (6) is:

[0025]

[0026] in Represents the derivative of x with respect to time, the state variable vector x, the control variable vector u and the matrix are defined as follows:

[0027] x=[P g1 P g2 P g P b Δf] T (7)

[0028] u=[u g u b ] T (8)

[0029]

[0030] Among them, P g1 and P g2 is the selected state variable, P g and P b are the output power of pumped storage and battery energy storage, Δf represents the frequency deviation of the power system, and u g and u b are the control signals for pumped storage and battery energy storage, T gPS is the governor constant of pumped storage in power generation mode, T R is the reset time, R T is the transient droop parameter, R PS is the static droop parameter, T w is the water starting time constant, D and M represent the load damping constant and inertia constant respectively; K B and T Brepresent the charge / discharge coefficient and time constant respectively, ACE is the area control error, and β represents the frequency deviation factor.

[0031] The state space equation model is discretized to obtain a discrete prediction model (12), which is specifically:

[0032] The forward Euler method (Formard-Euler method) is used to discretize the state space equation model (6). The discretization formula is shown in formula (11):

[0033]

[0034] Substituting equation (11) into the state space equation model (6), we get the discrete prediction model, which is described as follows:

[0035]

[0036] Where I represents the identity matrix and T is the sampling period.

[0037] The objective optimization function is:

[0038]

[0039] The constraints are:

[0040]

[0041] Where Q is the weight matrix of frequency deviation in the target optimization function; R is the weight matrix of control input in the target optimization function, is the rated power generation capacity of pumped storage, P g and P b are the output power of pumped storage and battery energy storage, E PS is the maximum storage capacity of pumped storage, P ESS is the power capacity of the battery energy storage, E ESS is the energy capacity of the battery storage, and They represent the states of battery energy storage and pumped hydroelectric storage at time k, and They represent the states of battery energy storage and pumped storage at time k+1, Δf(k+i) represents the frequency deviation of the power system at time k+i, and u g (k+i) and u b (k+i) are the control signals of pumped storage and battery energy storage at time k+i, respectively, and Δk represents the time interval of each step.

[0042] At each current moment, the optimization model composed of the target optimization function (13) and the constraint condition (14) is solved to obtain the optimal decision sequence; the first step of the optimal decision sequence is taken as the optimal action control instruction of the hybrid pumped storage-battery energy storage system at the current moment, and is sent to the hybrid pumped storage-battery energy storage system to participate in frequency modulation, specifically:

[0043] At the current time k, the current state x(k) of the actual power grid is measured by a synchronous phasor measurement device, with the goal of minimizing the sum of squares of frequency deviations in the next N steps, and taking the current state x(k) as the initial optimization condition, the optimization model composed of the target optimization function (13) and the constraint condition (14) is solved by the optimization algorithm interior point method to obtain the optimal decision sequence for the next N steps; the first step in the optimal decision sequence is taken as the optimal action control instruction of the hybrid pumped storage-battery energy storage system at the kth time, and is issued to the hybrid pumped storage-battery energy storage system to participate in frequency modulation; after the execution is completed, when the time comes to the k+Δkth time, the real-time state at the k+Δkth time is remeasured, and the above steps are repeated.

[0044] The present invention also provides a model predictive control device taking into account the coordinated frequency modulation of pumped storage and battery energy storage, comprising:

[0045] Dynamic model building module, used to build a dynamic model of hybrid pumped storage-battery energy storage system participating in secondary frequency regulation;

[0046] A state space equation model building module is used to build a state space equation model (6) equivalent to the dynamic model of the hybrid pumped storage-battery energy storage system participating in secondary frequency regulation;

[0047] A model discretization module, used for discretizing the state space equation model (6) to obtain a discrete prediction model (12);

[0048] An objective function establishment module is used to establish an objective optimization function (13) for minimizing the sum of squares of system frequency deviations according to a discrete prediction model (12), and to determine constraint conditions (14);

[0049] The frequency modulation control module is used to solve the optimization model composed of the target optimization function (13) and the constraint condition (14) at each current moment to obtain the optimal decision sequence; take the first step of the optimal decision sequence as the optimal action control instruction of the hybrid pumped storage-battery energy storage system at the current moment, and send it to the hybrid pumped storage-battery energy storage system to participate in frequency modulation.

[0050] The dynamic model of the hybrid pumped storage-battery energy storage system participating in secondary frequency regulation includes: a transfer function between turbine valve opening and mechanical power, a transfer function of the turbine speed governor, and a battery energy storage transfer function; wherein,

[0051] The transfer function between turbine valve opening and mechanical power is:

[0052]

[0053] Among them, ΔP m represents the output power deviation of pumped storage, ΔG is regarded as the gate opening deviation, s is the Laplace operator, T w is the water starting time constant, which is in the range between 0.5 s and 4.0 s at full load;

[0054] The transfer function of the turbine speed governor is:

[0055]

[0056] Where Δω represents the angular velocity deviation, T gPS is the governor constant of pumped storage in power generation mode, T R is the reset time, R T is the transient droop parameter, R PS is the static droop parameter;

[0057] The battery energy storage transfer function is:

[0058]

[0059] ΔP nL (s)-ΔP m (s)-ΔP B (s) = Δf(s)(D + sM) (4)

[0060] ΔP nL (s) = ΔP Load (s)-ΔP PV (s)-ΔP wind (s) (5)

[0061] Equations (4) and (5) represent the mathematical relationship between frequency deviation and unbalanced power. In the above equations, ΔP B is the output power deviation of the battery storage, u B Indicates automatic power generation control signal; K B and T B Denote the charge / discharge coefficient and time constant respectively; Δf denotes the frequency deviation of the power system; D and M denote the load damping constant and inertia constant respectively; ΔP nL is the net power change; ΔP Load , ΔP PV and ΔP wind They represent the changes in load demand, photovoltaic output power and wind power generation power respectively.

[0062] The state space equation model (6) is:

[0063]

[0064] in Represents the derivative of x with respect to time, the state variable vector x, the control variable vector u and the matrix are defined as follows:

[0065] x=[P g1 P g2 P g P b Δf] T (7)

[0066] u=[u g u b ] T (8)

[0067]

[0068]

[0069] Among them, P g1 and P g2 is the selected state variable, P g and P b are the output power of pumped storage and battery energy storage, Δf represents the frequency deviation of the power system, and u g and u b are the control signals for pumped storage and battery energy storage, T gPS is the governor constant of pumped storage in power generation mode, T R is the reset time, R T is the transient droop parameter, R PS is the static droop parameter, T w is the water starting time constant, D and M represent the load damping constant and inertia constant respectively; K B and T B represent the charge / discharge coefficient and time constant respectively, ACE is the area control error, and β represents the frequency deviation factor.

[0070] The model discrete module is specifically used for:

[0071] The forward Euler method is used to discretize the state space equation model (6). The discretization formula is shown in formula (11):

[0072]

[0073] Substituting equation (11) into the state space equation model (6), we get the discrete prediction model, which is described as follows:

[0074]

[0075] Where I represents the identity matrix and T is the sampling period.

[0076] The objective optimization function is:

[0077]

[0078] The constraints are:

[0079]

[0080] Where Q is the weight matrix of frequency deviation in the target optimization function; R is the weight matrix of control input in the target optimization function, P g rated is the rated power generation of pumped storage, P g and P b are the output power of pumped storage and battery energy storage, E PS is the maximum storage capacity of pumped storage, P ESS is the power capacity of the battery energy storage, E ESS is the energy capacity of the battery storage, and They represent the states of battery energy storage and pumped hydroelectric storage at time k, and They represent the states of battery energy storage and pumped storage at time k+1, Δf(k+i) represents the frequency deviation of the power system at time k+i, and u g (k+i) and u b (k+i) are the control signals of pumped storage and battery energy storage at time k+i, respectively, and Δk represents the time interval of each step.

[0081] The frequency modulation control module is specifically used for:

[0082] At the current time k, the current state x(k) of the actual power grid is measured by a synchronous phasor measurement device, with the goal of minimizing the sum of squares of frequency deviations in the next N steps, and taking the current state x(k) as the initial optimization condition, the optimization model composed of the target optimization function (13) and the constraint condition (14) is solved by the optimization algorithm interior point method to obtain the optimal decision sequence for the next N steps; the first step in the optimal decision sequence is taken as the optimal action control instruction of the hybrid pumped storage-battery energy storage system at the kth time, and is issued to the hybrid pumped storage-battery energy storage system to participate in frequency modulation; after the execution is completed, when the time comes to the k+Δkth time, the real-time state at the k+Δkth time is remeasured, and the above steps are repeated.

[0083] On the other hand, the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the model predictive control method for coordinated frequency modulation taking into account pumped storage and battery energy storage as described in the first aspect is implemented.

[0084] On the other hand, the present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the model predictive control method for coordinated frequency modulation taking into account pumped storage and battery energy storage as described in the first aspect.

[0085] The beneficial effects of the present invention are:

[0086] 1) The present invention considers the operating constraints of pumped storage and battery energy storage while determining the control signal, so that the charging and discharging power and the state of charge of the hybrid energy storage can always be kept within a safe range during the process of participating in the load frequency control of the power system.

[0087] 2) Different from the controller design method for single-input and single-output control objects in classical control theory, the present invention adopts the model prediction concept to achieve the design of control methods for multi-input objects based on the multi-input signal characteristics of hybrid energy storage control objects.

[0088] 3) The control strategy proposed in this invention provides necessary support for pumped storage-battery energy storage to participate in frequency control. In the case of wind and solar power fluctuations, the output of pumped storage and battery energy storage can be reasonably adjusted, and the frequency deviation can be adjusted to zero in a relatively short time. BRIEF DESCRIPTION OF THE DRAWINGS

[0089] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0090] Figure 1 It is a single-region power system frequency response model taking into account pumped storage and battery energy storage.

[0091] Figure 2 It is a control flow chart of pumped storage-battery energy storage based on model predictive control.

[0092] Figure 3 It is the frequency adjustment process after a power disturbance occurs.

[0093] Figure 4 It is the output power of battery energy storage during frequency regulation.

[0094] Figure 5 It is the output power of pumped storage during frequency regulation. DETAILED DESCRIPTION

[0095] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0096] The first aspect of the present invention provides a model predictive control method for coordinated frequency regulation of pumped storage and battery energy storage, which is a control method for the joint participation of pumped storage and battery energy storage in frequency regulation, and realizes the adjustment of power system frequency deviation when renewable energy power fluctuates.

[0097] This solution has the following features:

[0098] 1) A mathematical model of a hybrid pumped storage-battery energy storage system was established. First, a model was built based on the dynamic characteristics of pumped storage in power generation mode to obtain the transfer function relationship between the guide vane opening and the output mechanical power of the pumped storage. Secondly, the battery energy storage was modeled as a first-order inertia link to reflect the dynamic characteristics of the frequency modulation command and the output power. Finally, a steady-state operation model of the energy storage system was established.

[0099] 2) Considering the joint participation of pumped storage and battery energy storage in frequency control, a control strategy based on model prediction is designed. This control strategy has the following characteristics: first, model predictive control is suitable for such multi-input control objects; second, compared with the traditional proportional integral regulator, model predictive control fully considers the operating constraints of pumped storage and battery energy storage when determining the control signal.

[0100] According to the different dynamic working characteristics of pumped storage and battery energy storage, the present invention establishes corresponding mathematical models in the complex frequency domain, respectively, to achieve complementary utilization of the frequency regulation capability of hybrid energy storage.

[0101] The specific implementation methods are as follows:

[0102] A. Dynamic model of hybrid pumped storage-battery energy storage participating in load frequency control (secondary frequency regulation)

[0103] Firstly, a dynamic model of pumped storage participating in active power regulation under power generation state is established, and the transfer function between valve opening of the turbine and mechanical power can be expressed as:

[0104]

[0105] Among them, ΔP m represents the output power deviation of pumped storage, ΔG is regarded as the gate opening deviation, s is the Laplace operator, T w is the water start time constant, which is in the range between 0.5s and 4.0s at full load. Equation (1) describes the variation of mechanical output power with gate opening.

[0106] In order to ensure stable control performance, turbine control usually needs to add a transient droop link. Therefore, the transfer function of the speed governor can be selected as:

[0107]

[0108] Where Δω represents the angular velocity deviation, T gPS is the governor constant of pumped storage in power generation mode, T R is the reset time, R T is the transient droop parameter, R PS is the static droop parameter.

[0109] Secondly, the dynamic relationship of battery energy storage output power can be modeled by a first-order transfer function:

[0110]

[0111] Where ΔP B is the output power deviation of the battery storage, u B Indicates automatic power generation control signal. K B and T B represent the charge / discharge coefficient and time constant respectively.

[0112] The mathematical relationship between frequency deviation and unbalanced power can be described as:

[0113] ΔP nL (s)-ΔP m (s)-ΔP B (s) = Δf(s)(D + sM) (4)

[0114] ΔP nL (s) = ΔP Load (s)-ΔP PV (s)-ΔP wind (s) (5)

[0115] Where Δf represents the frequency deviation of the power system. D and M represent the load damping constant and inertia constant respectively. ΔP nL is the net power change. ΔP Load , ΔP PV and ΔP windThey represent the changes in load demand, photovoltaic output power and wind power generation power respectively.

[0116] B. Model predictive control method considering coordinated frequency regulation of pumped storage and battery energy storage

[0117] According to the above formulas (1)-(5), an equivalent state space equation model is established as the prediction model embedded in the controller:

[0118]

[0119] The state variable vector x, control variable vector u and matrix are defined as follows:

[0120] x=[P g1 P g2 P g P b Δf] T (7)

[0121] u=[u g u b ] T (8)

[0122]

[0123] Among them, P g1 and P g2 is the selected state variable, P g and P b are the output power of pumped storage and battery energy storage, u g and u b are the control signals for pumped storage and battery energy storage, ACE is the area control error, and β represents the frequency deviation factor.

[0124] In order to incorporate the state equation into the optimization of the control signal, the Formard-Euler method is used to discretize the continuous state space equation (6). The discrete formula is shown in (11):

[0125]

[0126] Substituting (11) into (6), we get the discrete prediction model, which is described as follows:

[0127]

[0128] Where I represents the identity matrix and T is the sampling period.

[0129] Secondly, model predictive control aims to determine the control signal u to ensure that the system output ACE is as close to zero as possible, which can be expressed as the following optimization problem:

[0130]

[0131] The constraints are as follows:

[0132]

[0133] Where Q is the weight matrix of frequency deviation in the target optimization function; R is the weight matrix of control input in the target optimization function, is the rated power generation capacity of pumped storage, E PS is the maximum storage capacity of pumped storage, P ESS is the power capacity of the battery energy storage, E ESS is the energy capacity of the battery storage, and They represent the states of battery energy storage and pumped hydroelectric storage at time k, and They represent the states of battery energy storage and pumped storage at time k+1, Δf(k+i) represents the frequency deviation of the power system at time k+i, and u g (k+i) and u b (k+i) are the control signals of pumped storage and battery energy storage at time k+i, respectively, and Δk represents the time interval of each step.

[0134] Subsequently, the first step of the optimal solution of the optimization model composed of (13) and (14) is applied to the actual frequency response model (15) considering random power fluctuations. This model is based on the prediction model (6) and further adds the random disturbance term ΔP to simulate the power fluctuation in the actual power grid secondary frequency regulation scenario. nl The result is shown in (15). Its function is to accept the frequency modulation action command of the model predictive controller based on the prediction model (12) and transmit the actual frequency result back to the controller as a feedback signal as the initial state of the optimization problem in the next step (time k+Δk).

[0135] Thus, at time k, pumped storage and battery storage respond to the MPC control signal and charge / discharge to regulate the frequency.

[0136]

[0137] At the k+Δkth time, the state of the model is remeasured or estimated as a feedback signal and the above steps are repeated.

[0138] The control method has the characteristics of rolling operation. After calculating the optimal decision sequence of the next 10 steps at time k, the decision result of the first step is selected for execution. After the execution is completed, it comes to time k+Δk and the above steps are repeated.

[0139] Figure 1The frequency response model of pumped storage and battery energy storage participating in the secondary frequency regulation of a single-region power system is described. The mathematical model is based on the transfer function module and describes the dynamic characteristics of pumped storage and battery energy storage, the inertia response of the power system, the primary frequency regulation, and the dynamic model of the secondary frequency regulation.

[0140] Figure 2 The model predictive control method is described in the process of hybrid pumped storage-battery energy storage participating in secondary frequency regulation. At the kth moment, the current state x(k) is first measured by the PMU device (synchronous phasor measurement device), and the measurement result will be used as the initial state of the controller recursive prediction (which can also be understood as the initial parameter condition of the optimization problem). Based on this, with the goal of minimizing the sum of squares of frequency deviations in the next N steps (13), the convex quadratic programming problem consisting of the objective optimization function (13) and the constraint condition (14) is solved by the interior point method to obtain the optimal decision sequence for the next N steps (i.e., the frequency regulation instructions of pumped storage-battery energy storage). Here, N is 10 steps, which is mainly determined by the actual engineering situation of secondary frequency regulation of the power grid. In practice, the frequency regulation instructions generally change once every 4 seconds, and the transition time is generally 5-10 seconds. The prediction time domain of the controller needs to be less than the above time to ensure the effectiveness of the prediction. According to experience, it is 10%-20%, so the prediction time domain can be taken as 1s. Considering that the controller time step (discrete step length) is 0.1s, N = 1 / 0.1 = 10 steps. The first step in the decision sequence is taken as the frequency modulation instruction to be finally executed at the kth moment and sent to the energy storage power station. After the execution is completed, when the time comes to the k+Δkth moment, the real-time state at the k+Δkth moment is re-measured and the above steps are repeated.

[0141] Based on the actual data of Tiantang Pumped Storage Power Station in Luotian County, Huanggang City, the proposed model predictive control method applied to pumped storage-new energy storage is verified below.

[0142] like Figure 3 As shown in Figure 1, the net load is reduced by 5kW due to the wind power / PV power fluctuation at the 10th second. In order to compensate for the unbalanced power within a few seconds, the proposed controller generates a control signal to the pumped storage / battery energy storage. Figure 4 As can be seen from the figure, the battery starts charging immediately at the 10th second to restore the frequency. Figure 5 As shown in Figure 1, the dynamic characteristics of pumped storage are relatively slow, and it enters the discharge mode to start generating electricity at the 12th second. From the 12th second, battery energy storage and pumped storage act as power sources together to alleviate the frequency deviation. Around the 26th second, the frequency almost recovers to 50Hz, which verifies the steady-state error performance of the proposed control method.

[0143] The embodiment of the present invention further provides a model predictive control device taking into account the coordinated frequency modulation of pumped storage and battery energy storage, comprising:

[0144] The dynamic model building module is used to build a dynamic model of the hybrid pumped storage-battery energy storage system participating in secondary frequency regulation.

[0145] A state space equation model building module is used to build a state space equation model (6) equivalent to the dynamic model of the hybrid pumped storage-battery energy storage system participating in secondary frequency regulation;

[0146] A model discretization module, used for discretizing the state space equation model (6) to obtain a discrete prediction model (12);

[0147] An objective function establishment module is used to establish an objective optimization function (13) for minimizing the sum of squares of system frequency deviations according to a discrete prediction model (12), and to determine constraint conditions (14);

[0148] The frequency modulation control module is used to solve the optimization model composed of the target optimization function (13) and the constraint condition (14) at each current moment to obtain the optimal decision sequence; take the first step of the optimal decision sequence as the optimal action control instruction of the hybrid pumped storage-battery energy storage system at the current moment, and send it to the hybrid pumped storage-battery energy storage system to participate in frequency modulation.

[0149] The dynamic model of the hybrid pumped storage-battery energy storage system participating in secondary frequency regulation includes: a transfer function between turbine valve opening and mechanical power, a transfer function of the turbine speed governor, and a battery energy storage transfer function; wherein,

[0150] The transfer function between turbine valve opening and mechanical power is:

[0151]

[0152] Where ΔP m represents the output power deviation of pumped storage, ΔG is regarded as the gate opening deviation, s is the Laplace operator, T w is the water starting time constant, which is in the range between 0.5 s and 4.0 s at full load;

[0153] The transfer function of the turbine speed governor is:

[0154]

[0155] Where Δω represents the angular velocity deviation, T gPS is the governor constant of pumped storage in power generation mode, T R is the reset time, R T is the transient droop parameter, R PS is the static droop parameter;

[0156] The battery energy storage transfer function is:

[0157]

[0158] ΔP nL (s)-ΔP m (s)-ΔP B (s) = Δf(s)(D + sM) (4)

[0159] ΔP nL (s) = ΔP Load (s)-ΔP PV (s)-ΔP wind (s) (5)

[0160] Equations (4) and (5) represent the mathematical relationship between frequency deviation and unbalanced power. In the above equations, ΔP B is the output power deviation of the battery storage, u B Indicates automatic power generation control signal; K B and T B Denote the charge / discharge coefficient and time constant respectively; Δf denotes the frequency deviation of the power system; D and M denote the load damping constant and inertia constant respectively; ΔP nL is the net power change; ΔP Load , ΔP PV and ΔP wind They represent the changes in load demand, photovoltaic output power and wind power generation power respectively.

[0161] The state space equation model (6) is:

[0162]

[0163] in Represents the derivative of x with respect to time, the state variable vector x, the control variable vector u and the matrix are defined as follows:

[0164] x=[P g1 P g2 P g P b Δf] T (7)

[0165] u=[u g u b ] T (8)

[0166]

[0167] Among them, P g1 and P g2 is the selected state variable, P g and P bare the output power of pumped storage and battery energy storage, Δf represents the frequency deviation of the power system, and u g and u b are the control signals for pumped storage and battery energy storage, T gPS is the governor constant of pumped storage in power generation mode, T R is the reset time, R T is the transient droop parameter, R PS is the static droop parameter, T w is the water starting time constant, D and M represent the load damping constant and inertia constant respectively; K B and T B represent the charge / discharge coefficient and time constant respectively, ACE is the area control error, and β represents the frequency deviation factor.

[0168] The model discrete module is specifically used for:

[0169] The forward Euler method is used to discretize the state space equation model (6). The discretization formula is shown in formula (11):

[0170]

[0171] Substituting equation (11) into the state space equation model (6), we get the discrete prediction model, which is described as follows:

[0172]

[0173] Where I represents the identity matrix and T is the sampling period.

[0174] The objective optimization function is:

[0175]

[0176] The constraints are:

[0177]

[0178] Where Q is the weight matrix of frequency deviation in the target optimization function; R is the weight matrix of control input in the target optimization function, P g rated is the rated power generation capacity of pumped storage, P g and P b are the output power of pumped storage and battery energy storage, E PS is the maximum storage capacity of pumped storage, P ESS is the power capacity of the battery energy storage, E ESS is the energy capacity of the battery storage, and They represent the states of battery energy storage and pumped hydroelectric storage at time k, and They represent the states of battery energy storage and pumped storage at time k+1, Δf(k+i) represents the frequency deviation of the power system at time k+i, and u g (k+i) and u b (k+i) are the control signals of pumped storage and battery energy storage at time k+i, respectively, and Δk represents the time interval of each step.

[0179] The frequency modulation control module is specifically used for:

[0180] At the current time k, the current state x(k) of the actual power grid is measured by a synchronous phasor measurement device, with the goal of minimizing the sum of squares of frequency deviations in the next N steps, and taking the current state x(k) as the initial optimization condition, the optimization model composed of the target optimization function (13) and the constraint condition (14) is solved by the optimization algorithm interior point method to obtain the optimal decision sequence for the next N steps; the first step in the optimal decision sequence is taken as the optimal action control instruction of the hybrid pumped storage-battery energy storage system at the kth time, and is sent to the hybrid pumped storage-battery energy storage system to participate in frequency modulation. After the execution is completed, when the time comes to the k+Δkth time, the real-time state at the k+Δkth time is remeasured, and the above steps are repeated.

[0181] On the other hand, the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the model predictive control method for coordinated frequency modulation taking into account pumped storage and battery energy storage as described in the first aspect is implemented.

[0182] On the other hand, the present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the model predictive control method for coordinated frequency modulation taking into account pumped storage and battery energy storage as described in the first aspect.

[0183] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0184] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0185] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0186] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0187] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A model predictive control method for coordinated frequency modulation of pumped storage and battery energy storage, characterized in that: include: Construct a dynamic model of a hybrid pumped storage-battery energy storage system participating in secondary frequency regulation; Based on the dynamic model of the hybrid pumped storage-battery energy storage system participating in secondary frequency regulation, an equivalent state space equation model (6) is established; Discretize the state space equation model (6) to obtain a discrete prediction model (12); According to the discrete prediction model (12), a target optimization function (13) is established to minimize the sum of squares of system frequency deviations, and constraint conditions (14) are determined; At each current moment, the optimization model composed of the objective optimization function (13) and the constraint condition (14) is solved to obtain the optimal decision sequence; the first step of the optimal decision sequence is taken as the optimal action control instruction of the hybrid pumped storage-battery energy storage system at the current moment, and is sent to the hybrid pumped storage-battery energy storage system to participate in frequency regulation.

2. The model predictive control method for coordinated frequency modulation of pumped storage and battery energy storage according to claim 1 is characterized in that: The dynamic model of the hybrid pumped storage-battery energy storage system participating in secondary frequency regulation includes: a transfer function between turbine valve opening and mechanical power, a transfer function of the turbine speed governor, and a battery energy storage transfer function; wherein, The transfer function between turbine valve opening and mechanical power is: Among them, ΔP m represents the output power deviation of pumped storage, ΔG is regarded as the gate opening deviation, s is the Laplace operator, T w is the water start-up time constant; The transfer function of the turbine speed governor is: Where Δω represents the angular velocity deviation, T gPS is the governor constant of pumped storage in power generation mode, T R is the reset time, R T is the transient droop parameter, R PS is the static droop parameter; The battery energy storage transfer function is: ΔP nL (s)-ΔP m (s)-ΔP B (s)=Δf(s)(D+sM) (4) ΔP nL (s)=ΔP Load (s)-ΔP PV (s)-ΔP wind (s) (5) Equations (4) and (5) represent the mathematical relationship between frequency deviation and unbalanced power. In the above equations, ΔP B is the output power deviation of the battery storage, u B Indicates automatic power generation control signal; K B and T B Denote the charge / discharge coefficient and time constant respectively; Δf denotes the frequency deviation of the power system; D and M denote the load damping constant and inertia constant respectively; ΔP nL is the net power change; ΔP Load , ΔP PV and ΔP wind They represent the changes in load demand, photovoltaic output power and wind power generation power respectively.

3. The model predictive control method for coordinated frequency modulation of pumped storage and battery energy storage according to claim 1 is characterized in that: The state space equation model (6) is: in Represents the derivative of x with respect to time, the state variable vector x, the control variable vector u and the matrix are defined as follows: x=[P g1 P g2 P g P b Δf] T (7) in=[in g in b ] T (8) Among them, P g1 and P g2 is the selected state variable, P g and P b are the output power of pumped storage and battery energy storage, Δf represents the frequency deviation of the power system, and u g and u b are the control signals for pumped storage and battery energy storage, T gPS is the governor constant of pumped storage in power generation mode, T R is the reset time, R T is the transient droop parameter, R PS is the static droop parameter, T w is the water starting time constant, D and M represent the load damping constant and inertia constant respectively; K B and T B represent the charge / discharge coefficient and time constant respectively, ACE is the area control error, and β represents the frequency deviation factor.

4. The model predictive control method for coordinated frequency modulation of pumped storage and battery energy storage according to claim 3 is characterized in that: The state space equation model is discretized to obtain a discrete prediction model (12), which is specifically: The forward Euler method is used to discretize the state space equation model (6). The discretization formula is shown in formula (11): Substituting equation (11) into the state space equation model (6), we get the discrete prediction model, which is described as follows: Where I represents the identity matrix and T is the sampling period.

5. The model predictive control method for coordinated frequency modulation of pumped storage and battery energy storage according to claim 1 is characterized in that: The objective optimization function is: The constraints are: Where Q is the weight matrix of frequency deviation in the target optimization function; R is the weight matrix of control input in the target optimization function, is the rated power generation capacity of pumped storage, P g and P b are the output power of pumped storage and battery energy storage, E PS is the maximum storage capacity of pumped storage, P ESS is the power capacity of the battery energy storage, E ESS is the energy capacity of the battery storage, and They represent the states of battery energy storage and pumped hydroelectric storage at time k, and They represent the states of battery energy storage and pumped storage at time k+1, Δf(k+i) represents the frequency deviation of the power system at time k+i, and u g (k+i) and u b (k+i) are the control signals of pumped storage and battery energy storage at time k+i, respectively, and Δk represents the time interval of each step.

6. The model predictive control method for coordinated frequency modulation of pumped storage and battery energy storage according to claim 1 is characterized in that: At each current moment, the optimization model composed of the target optimization function (13) and the constraint condition (14) is solved to obtain the optimal decision sequence; the first step of the optimal decision sequence is taken as the optimal action control instruction of the hybrid pumped storage-battery energy storage system at the current moment, and is sent to the hybrid pumped storage-battery energy storage system to participate in frequency modulation, specifically: At the current time k, the current state x(k) of the actual power grid is measured by a synchronous phasor measurement device, with the goal of minimizing the sum of squares of frequency deviations in the next N steps, and taking the current state x(k) as the initial optimization condition, the optimization model composed of the target optimization function (13) and the constraint condition (14) is solved by the optimization algorithm interior point method to obtain the optimal decision sequence for the next N steps; the first step in the optimal decision sequence is taken as the optimal action control instruction of the hybrid pumped storage-battery energy storage system at the kth time, and is sent to the hybrid pumped storage-battery energy storage system to participate in frequency regulation.

7. A model predictive control device taking into account the coordinated frequency modulation of pumped storage and battery energy storage, characterized in that: include: Dynamic model building module, used to build a dynamic model of hybrid pumped storage-battery energy storage system participating in secondary frequency regulation; A state space equation model building module is used to build a state space equation model (6) equivalent to the dynamic model of the hybrid pumped storage-battery energy storage system participating in secondary frequency regulation; A model discretization module, used for discretizing the state space equation model (6) to obtain a discrete prediction model (12); An objective function establishment module is used to establish an objective optimization function (13) for minimizing the sum of squares of system frequency deviations according to a discrete prediction model (12), and to determine constraint conditions (14); The frequency modulation control module is used to solve the optimization model composed of the target optimization function (13) and the constraint condition (14) at each current moment to obtain the optimal decision sequence; take the first step of the optimal decision sequence as the optimal action control instruction of the hybrid pumped storage-battery energy storage system at the current moment, and send it to the hybrid pumped storage-battery energy storage system to participate in frequency modulation.

8. The model predictive control device for coordinated frequency modulation taking into account pumped storage and battery energy storage according to claim 7, characterized in that: The dynamic model of the hybrid pumped storage-battery energy storage system participating in secondary frequency regulation includes: a transfer function between turbine valve opening and mechanical power, a transfer function of the turbine speed governor, and a battery energy storage transfer function; wherein, The transfer function between turbine valve opening and mechanical power is: Among them, ΔP m represents the output power deviation of pumped storage, ΔG is regarded as the gate opening deviation, s is the Laplace operator, T w is the water starting time constant, which is in the range between 0.5 s and 4.0 s at full load; The transfer function of the turbine speed governor is: Where Δω represents the angular velocity deviation, T gPS is the governor constant of pumped storage in power generation mode, T R is the reset time, R T is the transient droop parameter, R PS is the static droop parameter; The battery energy storage transfer function is: ΔP nL (s)-ΔP m (s)-ΔP B (s)=Δf(s)(D+sM) (4) ΔP nL (s)=ΔP Load (s)-ΔP PV (s)-ΔP wind (s) (5) Equations (4) and (5) represent the mathematical relationship between frequency deviation and unbalanced power. In the above equations, ΔP B is the output power deviation of the battery storage, u B Indicates automatic power generation control signal; K B and T B Denote the charge / discharge coefficient and time constant respectively; Δf denotes the frequency deviation of the power system; D and M denote the load damping constant and inertia constant respectively; ΔP nL is the net power change; ΔP Load , ΔP PV and ΔP wind They represent the changes in load demand, photovoltaic output power and wind power generation power respectively.

9. The model predictive control device for coordinated frequency modulation taking into account pumped storage and battery energy storage according to claim 7, characterized in that: The state space equation model (6) is: in Represents the derivative of x with respect to time, the state variable vector x, the control variable vector u and the matrix are defined as follows: x=[P g1 P g2 P g P b Δf] T (7) in=[in g in b ] T (8) Among them, P g1 and P g2 is the selected state variable, P g and P b are the output power of pumped storage and battery energy storage, Δf represents the frequency deviation of the power system, and u g and u b are the control signals for pumped storage and battery energy storage, T gPS is the governor constant of pumped storage in power generation mode, T R is the reset time, R T is the transient droop parameter, R PS is the static droop parameter, T w is the water starting time constant, D and M represent the load damping constant and inertia constant respectively; K B and T B represent the charge / discharge coefficient and time constant respectively, ACE is the area control error, and β represents the frequency deviation factor.

10. The model predictive control device for coordinated frequency modulation taking into account pumped storage and battery energy storage according to claim 7, characterized in that: The model discrete module is specifically used for: The forward Euler method is used to discretize the state space equation model (6). The discretization formula is shown in formula (11): Substituting equation (11) into the state space equation model (6), we get the discrete prediction model, which is described as follows: Where I represents the identity matrix and T is the sampling period.

11. The model predictive control device for coordinated frequency modulation taking into account pumped storage and battery energy storage according to claim 7, characterized in that: The objective optimization function is: The constraints are: Where Q is the weight matrix of frequency deviation in the target optimization function; R is the weight matrix of control input in the target optimization function, is the rated power generation capacity of pumped storage, P g and P b are the output power of pumped storage and battery energy storage, E PS is the maximum storage capacity of pumped storage, P ESS is the power capacity of the battery energy storage, E ESS is the energy capacity of the battery storage, and They represent the states of battery energy storage and pumped hydroelectric storage at time k, and They represent the states of battery energy storage and pumped storage at time k+1, Δf(k+i) represents the frequency deviation of the power system at time k+i, and u g (k+i) and u b (k+i) are the control signals of pumped storage and battery energy storage at time k+i, respectively, and Δk represents the time interval of each step.

12. The model predictive control device for coordinated frequency modulation taking into account pumped storage and battery energy storage according to claim 7, characterized in that: The frequency modulation control module is specifically used for: At the current time k, the current state x(k) of the actual power grid is measured by a synchronous phasor measurement device, with the goal of minimizing the sum of squares of frequency deviations in the next N steps, and taking the current state x(k) as the initial optimization condition, the optimization model composed of the target optimization function (13) and the constraint condition (14) is solved by the optimization algorithm interior point method to obtain the optimal decision sequence for the next N steps; the first step in the optimal decision sequence is taken as the optimal action control instruction of the hybrid pumped storage-battery energy storage system at the kth time, and is sent to the hybrid pumped storage-battery energy storage system to participate in frequency regulation.

13. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method for model predictive control taking into account the coordinated frequency modulation of pumped storage and battery energy storage as described in any one of claims 1 to 6 is implemented.

14. A non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the model predictive control method for coordinated frequency modulation of pumped storage and battery energy storage as described in any one of claims 1 to 6.

Images