State estimation method of power distribution network

Abstract

The invention relates to a state estimation method for a power distribution network, and belongs to the technical field of power system operation and control. According to the method, the prior state quantity and the measured value are substituted into the power distribution network state estimation model established based on the improved extended Kalman filtering to estimate the system node state of the power distribution network in real time, the prediction process is replaced, and the calculation complexity is reduced. According to the method, the observability partitioning method is adopted, dynamic real-time partitioning is carried out on the network based on measurement collection conditions at different moments, the real-time requirement of power distribution network estimation is met, the observable area state is mapped to the unobservable area, the situation of insufficient measurement redundancy is compensated on the basis that pseudo measurement is not carried out, and the real-time performance of power distribution network estimation is improved. Therefore, high-precision state estimation of the power distribution network is realized.

Description

Technical Field

[0001] The invention relates to a state estimation method for a distribution network, and belongs to the technical field of power system operation and control. Background Art

[0002] In recent years, as the energy structure transition has progressed, the increasing number of distributed power sources and energy storage systems connected to distribution networks has made them increasingly complex. Furthermore, the random nature of distributed power output causes frequent changes in grid operating conditions. To meet the power transmission and distribution needs of complex distribution networks, it is crucial to utilize state estimation to capture the real-time operating status of distribution networks.

[0003] Observability analysis is a prerequisite for state estimation. Observability means that a system is observable if the state determined by a finite number of observations (measurements) is unique. A system is observable if the values of all its state variables can be inferred from existing measurement data and network topology information. Observability analysis helps determine whether the available measurement data is comprehensive enough to ensure accurate state estimation. To meet system observability requirements, it is necessary to increase the system's measurement redundancy. Existing technologies generally achieve this through pseudo-measurement modeling. Pseudo-measurements combine the system's physical model with known measurement information to infer variables that cannot be directly measured. To a certain extent, pseudo-measurements can effectively supplement real-time measurements, especially when it is difficult to directly install sensors at measurement points, providing estimates of the system state. However, pseudo-measurements are not precise real-time data. Their calculation relies on mathematical models and assumptions, which can lead to errors due to incomplete models or inaccurate parameters. These errors can gradually accumulate during system state estimation, affecting the accuracy of the overall estimation. Summary of the Invention

[0004] The technical problem to be solved by the present invention is: how to quickly and accurately estimate the state of a distribution network.

[0005] To solve the above technical problems, the present invention proposes a technical solution: a state estimation method for a distribution network, comprising the following steps:

[0006] Step 1: Obtain the topology and network parameters of the distribution network for which state estimation is to be performed; the topology includes the number of nodes and branches of the distribution network, and the network parameters include the branch number, branch head node and terminal node number, branch conductance, and branch susceptance of the distribution network;

[0007] Set the first time t1 for the distribution network to be ready for the first state estimation, the second time t2 for the second state estimation, and the nth time t n ;

[0008] Step 2: Establish a distribution network state estimation model based on the improved extended Kalman filter algorithm, as shown in the following formula (1):

[0009]

[0010] In formula (1), is the predicted state quantity obtained by performing state estimation on the distribution network at the kth moment; is the prior state quantity of the distribution network at the kth moment; is the Kalman gain equation of the distribution network at the kth moment; is the measurement value collected in real time by the distribution network at the kth moment; is the predicted measurement function value of the distribution network at the kth moment; is the estimated error covariance matrix of the distribution network between the k-1th moment and the kth moment; and are the Jacobian matrix and its transpose of the distribution network at the kth moment respectively; R is the measurement noise covariance matrix when the measurement value of the distribution network is collected in real time; is the predicted initial value of the state quantity of the distribution network at the kth moment, which is an empirical value; Step 3: real-time acquisition of the real-time node voltage amplitude, real-time node injected active power, real-time node injected reactive power of n nodes of the distribution network within the real-time initial moment t0 closest to the first moment t1 and the real-time active power and real-time reactive power of the branch head end of m branches and collect them to form the initial measurement value of the distribution network at the real-time initial moment t0 As shown in the following formula (2),

[0011]

[0012] In formula (2), is the real-time node voltage amplitude from the first node to the nth node among the n nodes of the distribution network within the real-time initial time t0; The real-time node active power injected from the first node to the n-th node among the n nodes of the distribution network within the real-time initial time t0; The reactive power is injected from the first node to the nth node of the distribution network within the real-time initial time t0; is the real-time active power of the first end of the branch to the m-th branch of the distribution network within the real-time initial time t0; is the real-time reactive power of the first end of the branch from the first branch to the m-th branch of the distribution network within the real-time initial time t0;

[0013] The initial predicted state quantity of the distribution network at the real-time initial time t0 is obtained by weighted least square method As shown in the following formula (3),

[0014]

[0015] In formula (3), is the predicted node voltage amplitude from the first node to the nth node among the n nodes of the distribution network within the real-time initial time t0; is the predicted voltage phase angle from the first node to the nth node among the n nodes of the distribution network within the real-time initial time t0;

[0016] Step 4: The initial predicted state quantity As the first a priori state quantity of the distribution network at the first time t1 And put it into the following formula (4) to calculate the first measurement function of the distribution network at the first moment t1: and the first Jacobian matrix

[0017]

[0018] In formula (4), and The active power and reactive power are injected into the prediction nodes from the first node to the nth node among the n nodes of the distribution network within the real-time initial time t0;

[0019] and The active power and reactive power are injected into the prediction nodes of the first branch to the mth branch of the distribution network within the real-time initial time t0 respectively; G ii+1 and B ii+1 are the conductance and susceptance between the i-th node and the i+1-th node among the n nodes of the distribution network, respectively; is the predicted voltage phase angle difference between the i-th node and the i+1-th node among the n nodes of the distribution network; and are respectively the predicted voltage amplitude at the head end and the predicted voltage amplitude at the tail end of the j-th branch among the m branches of the distribution network at the real-time initial time t0; G j and B j are the conductance and susceptance of the j-th branch among the m branches of the distribution network, respectively; is the predicted voltage phase angle difference between the head node and the tail node of the j-th branch among the m branches of the distribution network;

[0020] Set the initial value of the first predicted state quantity of the distribution network at the first time t1 and the measurement noise covariance matrix; real-time acquisition of the real-time node voltage amplitude, real-time node injected active power, real-time node injected reactive power of the n nodes of the distribution network within the first time t1 and the real-time active power and real-time reactive power of the branch head end of the m branches and collect them to form a first measurement value The first prior state quantity First measurement function First Jacobian matrix The initial value of the first predicted state quantity and the first measurement value Substitute into the distribution network state estimation model to obtain the first predicted state quantity of the distribution network at the first time t1 As shown in the following formula (5),

[0021]

[0022] In formula (5), is the predicted node voltage amplitude from the first node to the nth node among the n nodes of the distribution network within the first time t1; is the predicted voltage phase angle from the first node to the nth node among the n nodes of the distribution network within the first time t1;

[0023] The first predicted state quantity All predicted node voltage amplitudes and phase angles in are collected to form the first predicted node voltage phasor sequence The first predicted node voltage phasor sequence As the first state estimation result of the distribution network at the first time t1

[0024] Step 5: The first predicted state quantity As the second priori state quantity of the distribution network at the second time t2 And according to the principle of formula (4), the second measurement function of the distribution network at the second moment t2 is calculated as follows: and the second Jacobian matrix

[0025] Setting the initial value of the second predicted state quantity of the distribution network at the second time t2 and measuring the noise covariance matrix; collecting the real-time node voltage amplitude, real-time node injected active power, real-time node injected reactive power of the n nodes of the distribution network at the second time t2 and the real-time active power and real-time reactive power of the branch head end of the m branches in real time and collecting them to form a second measurement value The second prior state quantity Second measurement function Second Jacobian matrix The initial value of the second predicted state quantity and the second measurement value Substitute into the distribution network state estimation model to obtain the second predicted state quantity of the distribution network at the second time t2 As shown in the following formula (6),

[0026]

[0027] In formula (6), is the predicted node voltage amplitude from the first node to the nth node among the n nodes of the distribution network within the second time t2; is the predicted voltage phase angle from the first node to the nth node among the n nodes of the distribution network within the second time t2;

[0028] The second predicted state quantity All predicted node voltage amplitudes and phase angles in are collected to form the second predicted node voltage phasor sequence The second predicted node voltage phasor sequence As the second state estimation result of the distribution network at the second time t2 Step 6: Repeat the principle of step 5 to obtain the distribution network from the third time t3 to the nth time t n The state estimation result of the distribution network is completed in the time period [t1,t n ] n state estimates within ].

[0029] Furthermore, the step 4 further includes the following:

[0030] The observability regions of n nodes in the distribution network are divided into node state observable regions Node ob and node status unobservable area Node un ; The first state estimation result Node in the node status observable area ob The predicted node voltage amplitudes and phase angles of the nodes in the are collected to form the first observable region voltage phasor prediction sequence U ob1 ; The first state estimation result The node status belongs to the unobservable area Node un The predicted node voltage amplitudes and phase angles of the nodes in the are collected to form the first unobservable region voltage phasor prediction sequence U un1 ;

[0031] Establish the node status observable area Node un and node status unobservable area Node ob The state space mapping model between is shown in the following formula (7):

[0032] In formula (7), U′ un is the voltage phasor correction prediction sequence of the unobservable area; H is the Jacobian matrix; U ob is the voltage phasor prediction sequence of the observable area; Y un is the admittance matrix of the unobservable region; Y L is a diagonal matrix whose diagonal elements are unobservable node admittances; Y u is the unobservable node admittance sequence; P un and Q un is the active power and reactive power of the unobservable node; U un Prediction sequence of voltage phasors for unobservable regions;

[0033] The first observable area voltage phasor prediction sequence U ob1 and the first unobservable region voltage phasor prediction sequence U un1 Substitute into the state space mapping model to calculate the first unobservable area voltage phasor correction prediction sequence U' of the distribution network at the first time t1 un1 The first observable regional voltage phasor prediction sequence U ob1 and the first unobservable region voltage phasor correction prediction sequence U′ un1 Collected to form the first revised state estimation result of the distribution network at the first time t1

[0034] Furthermore, the step 5 further includes the following:

[0035] The second state estimation result Node in the node status observable area ob The predicted node voltage amplitudes and phase angles of the nodes in the are collected to form the second observable area voltage phasor prediction sequence U ob2 ; The second state estimation result The node status belongs to the unobservable area Node unThe predicted node voltage amplitudes and phase angles of the nodes in the are collected to form the second unobservable region voltage phasor prediction sequence U un2 ;

[0036] The second observable area voltage phasor prediction sequence U ob2 And the second unobservable region voltage phasor prediction sequence U un2 Substitute into the state space mapping model to calculate the second unobservable area voltage phasor correction prediction sequence U′ of the distribution network at the second time t2 un2 The second observable area voltage phasor prediction sequence U ob2 and the second unobservable region voltage phasor correction prediction sequence U′ un2 Collected to form the second revised state estimation result of the distribution network at the second time t2

[0037] The beneficial effects of the present invention are as follows: 1. The present invention performs real-time estimation of the system node states of the distribution network by substituting prior state quantities and measurement values into a distribution network state estimation model established based on an improved extended Kalman filter, replacing the prediction process and reducing computational complexity. 2. The present invention adopts an observability partitioning method to dynamically partition the network in real time based on measurement collection conditions at different times, meeting the real-time requirements of distribution network estimation, and mapping the observable area states to the unobservable areas, compensating for the lack of measurement redundancy without performing pseudo-measurements, thereby achieving high-precision distribution network state estimation. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 This is a flow chart of a method for estimating the state of a distribution network in the first embodiment of the present invention.

[0039] Figure 2 This is a flow chart of a method for estimating the state of a distribution network in the second embodiment of the present invention.

[0040] Figure 3 It is a comparison diagram of the state estimation results of the present invention. DETAILED DESCRIPTION

[0041] The following further describes a method for estimating the state of a distribution network according to the present invention in conjunction with the accompanying drawings and specific embodiments.

[0042] Example 1

[0043] The state estimation method of the distribution network in this embodiment is as follows: Figure 1 As shown, the following steps are included:

[0044] Step 1: Obtain the topology and network parameters of the distribution network for state estimation. The topology includes the number of nodes and branches in the distribution network, and the network parameters include the branch number, the branch head node and terminal node numbers, branch conductance, and branch susceptance.

[0045] Set the first time t1 for the distribution network to be ready for the first state estimation, the second time t2 for the second state estimation, and the nth time t n ;

[0046] Step 2: Establish a distribution network state estimation model based on the improved extended Kalman filter algorithm, as shown in the following formula (1):

[0047]

[0048] In formula (1), It is the predicted state quantity obtained by state estimation of the distribution network at the kth moment; is the prior state quantity of the distribution network at the kth moment; is the Kalman gain equation of the distribution network at the kth moment; is the measurement value collected in real time by the distribution network at the kth moment; is the predicted measurement function value of the distribution network at the kth moment; is the estimated error covariance matrix of the distribution network from the k-1th moment to the kth moment; and are the Jacobian matrix and its transpose of the distribution network at the kth moment; R is the measurement noise covariance matrix when the measurement values of the distribution network are collected in real time; is the initial value of the predicted state quantity of the distribution network at the kth moment, which is an empirical value;

[0049] Step 3: Real-time collection of the real-time node voltage amplitude, real-time node injected active power, real-time node injected reactive power of n nodes of the distribution network within the real-time initial time t0 closest to the first time t1 and the real-time active power and real-time reactive power of the branch head end of m branches and collect them to form the initial measurement value of the distribution network within the real-time initial time t0 As shown in the following formula (2),

[0050]

[0051] In formula (2), is the real-time node voltage amplitude from the first node to the nth node among the n nodes of the distribution network at the real-time initial time t0; It is the real-time node injected active power from the first node to the nth node among the n nodes of the distribution network within the real-time initial time t0; It is the real-time node injected reactive power from the first node to the nth node among the n nodes of the distribution network within the real-time initial time t0; is the real-time active power at the branch head end from the first branch to the m-th branch of the distribution network within the real-time initial time t0; is the real-time reactive power at the branch head end from the first branch to the m-th branch of the distribution network within the real-time initial time t0;

[0052] The initial predicted state quantity of the distribution network at the real-time initial time t0 is obtained by using the existing technology weighted least squares method (WLS), the specific source of which can be found in the Chinese invention patent application document with application publication number CN103972884A entitled "A method for estimating the state of a power system" and paragraphs 26 to 27. As shown in the following formula (3),

[0053]

[0054] In formula (3), is the predicted node voltage amplitude from the first node to the nth node in the distribution network at the real-time initial time t0; is the predicted voltage phase angle from the first node to the nth node in the distribution network within the real-time initial time t0;

[0055] Step 4: Initial prediction state As the first a priori state quantity of the distribution network at the first moment t1 And put it into the following formula (4) to calculate the first measurement function of the distribution network at the first moment t1: and the first Jacobian matrix

[0056]

[0057] In formula (4), and They are respectively the predicted node injected active power and reactive power from the first node to the nth node in the distribution network within the real-time initial time t0; and are the active power and reactive power injected by the prediction nodes from the first branch to the mth branch of the distribution network within the real-time initial time t0; G ii+1 and B ii+1 are the conductance and susceptance between the i-th node and the i+1-th node in the n nodes of the distribution network; is the predicted voltage phase angle difference between the i-th node and the i+1-th node among the n nodes of the distribution network; and are the predicted voltage amplitude at the head end and the predicted voltage amplitude at the tail end of the j-th branch in the m branches of the distribution network at the real-time initial time t0; G j and B j are the conductance and susceptance of the jth branch among the m branches of the distribution network; is the predicted voltage phase angle difference between the head node and the tail node of the j-th branch among the m branches of the distribution network;

[0058] Set the initial value of the first predicted state quantity of the distribution network at the first time t1 and measurement noise covariance matrix; real-time node voltage amplitude, real-time node injected active power, real-time node injected reactive power of n nodes of the distribution network within the first moment t1 and real-time active power and real-time reactive power of the branch head end of m branches are collected and formed into a first measurement value The first prior state quantity First measurement function First Jacobian matrix The initial value of the first predicted state quantity and the first measurement value Substitute into the distribution network state estimation model to obtain the first predicted state quantity of the distribution network at the first time t1 As shown in the following formula (5),

[0059]

[0060] In formula (5), is the predicted node voltage amplitude from the first node to the nth node in the distribution network at the first time t1; is the predicted voltage phase angle from the first node to the nth node in the distribution network at the first time t1;

[0061] The first predicted state quantity All predicted node voltage amplitudes and phase angles in are collected to form the first predicted node voltage phasor sequence The first predicted node voltage phasor sequence As the first state estimation result of the distribution network at the first time t1

[0062] Step 5: The first predicted state quantity As the second priori state quantity of the distribution network at the second time t2 And according to the principle of formula (4), the second measurement function of the distribution network at the second time t2 is calculated as and the second Jacobian matrix

[0063] Set the initial value of the second predicted state quantity of the distribution network at the second time t2 and measurement noise covariance matrix; real-time acquisition of the real-time node voltage amplitude, real-time node injected active power, real-time node injected reactive power of the n nodes of the distribution network at the second time t2 and the real-time active power and real-time reactive power of the branch head end of the m branches and collect them to form a second measurement value The second prior state quantity Second measurement function Second Jacobian matrix The initial value of the second predicted state quantity and the second measurement value Substitute into the distribution network state estimation model to obtain the second predicted state quantity of the distribution network at the second time t2 As shown in the following formula (6),

[0064]

[0065] In formula (6), is the predicted node voltage amplitude from the first node to the nth node among the n nodes of the distribution network at the second time t2; is the predicted voltage phase angle from the first node to the nth node among the n nodes of the distribution network at the second time t2;

[0066] The second predicted state quantity All predicted node voltage amplitudes and phase angles in are collected to form the second predicted node voltage phasor sequence The second predicted node voltage phasor sequence As the second state estimation result of the distribution network at the second time t2

[0067] Step 6: Repeat the principle of step 5 to obtain the distribution network from the third time t3 to the nth time t n The state estimation results of the distribution network are completed in the time period [t1,t n ] n state estimates within ].

[0068] Example 2

[0069] This embodiment further includes the following contents based on the first embodiment: Figure 3 As shown, where:

[0070] Step 4 also includes the following:

[0071] The observability regions of n nodes in the distribution network are divided into node state observable regions Node ob and node status unobservable area Nodeun ; The first state estimation result Node in the node status observable area ob The predicted node voltage amplitudes and phase angles of the nodes in the are collected to form the first observable region voltage phasor prediction sequence U ob1 ; The first state estimation result The node status belongs to the unobservable area Node un The predicted node voltage amplitudes and phase angles of the nodes in the are collected to form the first unobservable region voltage phasor prediction sequence U un1 ;

[0072] Establish node status observable area Node un and node status unobservable area Node ob The state space mapping model between is shown in the following formula (7):

[0073]

[0074] In formula (7), U′ un is the voltage phasor correction prediction sequence of the unobservable area; H is the Jacobian matrix; U ob is the voltage phasor prediction sequence of the observable area; Y un is the admittance matrix of the unobservable region; Y L is a diagonal matrix whose diagonal elements are unobservable node admittances; Y u is the unobservable node admittance sequence; P un and Q un is the active power and reactive power of the unobservable node; U un Prediction sequence of voltage phasors for unobservable regions;

[0075] The first observable area voltage phasor prediction sequence U ob1 and the first unobservable region voltage phasor prediction sequence U un1 Substitute into the state space mapping model to calculate the first unobservable area voltage phasor correction prediction sequence U' of the distribution network at the first time t1 un1 The first observable regional voltage phasor prediction sequence U ob1 and the first unobservable region voltage phasor correction prediction sequence U′ un1 Collected to form the first revised state estimation result of the distribution network at the first time t1

[0076] Step 5 also includes the following:

[0077] The second state estimation result Node in the node status observable area ob The predicted node voltage amplitudes and phase angles of the nodes in the are collected to form the second observable area voltage phasor prediction sequence U ob2 ; The second state estimation result The node status belongs to the unobservable area Node un The predicted node voltage amplitudes and phase angles of the nodes in the are collected to form the second unobservable region voltage phasor prediction sequence U un2 ;

[0078] The second observable area voltage phasor prediction sequence U ob2 And the second unobservable region voltage phasor prediction sequence U un2 Substitute into the state space mapping model to calculate the second unobservable area voltage phasor correction prediction sequence U′ of the distribution network at the second time t2 un2 The second observable area voltage phasor prediction sequence U ob2 and the second unobservable region voltage phasor correction prediction sequence U′ un2 Collected to form the second revised state estimation result of the distribution network at the second time t2

[0079] In order to verify the effectiveness of the method of the present invention, a 33-node three-phase balanced distribution network is used for verification and explanation below.

[0080] The distribution network measurement data is obtained by adding Gaussian noise to the true value of the power flow. The standard deviation of the node voltage amplitude measurement noise is 0.01 pu, and the standard deviation of the node injection power and branch head end power measurement noise is 0.001 pu. The present invention is tested in a specific measurement data collection scenario: 50% of the nodes have node voltage amplitude, node injection active power, and node injection reactive power, and 30% of the branches have branch head end active power and branch head end reactive power. This embodiment selects the WLS algorithm as the comparison algorithm. This algorithm performs pseudo-measurement completion for the missing node voltage amplitude, node injection active power, and node injection reactive power.

[0081] In order to facilitate quantitative analysis of the state assessment accuracy, the mean absolute error of voltage amplitude and the mean absolute error of phase angle can be used as measurement indicators:

[0082]

[0083] in, and x true They represent the estimated value and true value of the state quantity respectively, and the state quantity is the voltage amplitude.

[0084] The voltage amplitude estimation error of each algorithm according to the node number is as follows: Figure 3 As shown in Table 1, the estimation accuracy of the two algorithms in this specific measurement acquisition scenario is shown. It can be seen that the estimation accuracy of the algorithm proposed in this patent is higher than that of the WLS algorithm. When the number of measurements is insufficient, WLS uses pseudo-measurements to supplement the missing node voltage amplitude, node injected active power, and node injected reactive power. However, pseudo-measurements may have certain errors, resulting in the estimation accuracy failing to meet the requirements. The algorithm of the present invention has a better estimation effect.

[0085] Table 1 Estimation errors of different estimation methods

[0086]

[0087] The above simulation results verify the effectiveness and practicality of the method of the present invention. Therefore, the method of the present invention makes up for the disadvantage of low estimation accuracy when the measurement redundancy is low, does not require additional pseudo-measurements, and improves the state estimation accuracy.

Claims

1. A method for state estimation of a distribution network, characterized by: The following steps are involved: Step 1: Obtain the topology and network parameters of the distribution network for which state estimation is to be performed; the topology includes the number of nodes and branches of the distribution network, and the network parameters include the branch number, branch head node and terminal node number, branch conductance, and branch susceptance of the distribution network; Set the first time t1 for the distribution network to be ready for the first state estimation, the second time t2 for the second state estimation, and the nth time t n ; Step 2: Establish a distribution network state estimation model based on the improved extended Kalman filter algorithm, as shown in the following formula (1): In formula (1), is the predicted state quantity obtained by performing state estimation on the distribution network at the kth moment; is the prior state quantity of the distribution network at the kth moment; is the Kalman gain equation of the distribution network at the kth moment; is the measurement value collected in real time by the distribution network at the kth moment; is the predicted measurement function value of the distribution network at the kth moment; is the estimated error covariance matrix of the distribution network between the k-1th moment and the kth moment; and are the Jacobian matrix and its transpose of the distribution network at the kth moment respectively; R is the measurement noise covariance matrix when the measurement values of the distribution network are collected in real time; is the predicted initial value of the state quantity of the distribution network at the kth moment, which is an empirical value; Step 3: real-time acquisition of the real-time node voltage amplitude, real-time node injected active power, real-time node injected reactive power of n nodes of the distribution network within the real-time initial moment t0 closest to the first moment t1 and the real-time active power and real-time reactive power of the branch head end of m branches and collect them to form the initial measurement value of the distribution network at the real-time initial moment t0 As shown in the following formula (2), In formula (2), is the real-time node voltage amplitude from the first node to the nth node among the n nodes of the distribution network within the real-time initial time t0; The real-time node active power injected from the first node to the n-th node among the n nodes of the distribution network within the real-time initial time t0; The reactive power is injected from the first node to the nth node of the distribution network within the real-time initial time t0; is the real-time active power of the first end of the branch to the m-th branch of the distribution network within the real-time initial time t0; is the real-time reactive power of the first end of the branch from the first branch to the m-th branch of the distribution network within the real-time initial time t0; The initial predicted state quantity of the distribution network at the real-time initial time t0 is obtained by weighted least square method As shown in the following formula (3), In formula (3), is the predicted node voltage amplitude from the first node to the nth node among the n nodes of the distribution network within the real-time initial time t0; is the predicted voltage phase angle from the first node to the nth node among the n nodes of the distribution network within the real-time initial time t0; Step 4: The initial predicted state quantity As the first a priori state quantity of the distribution network at the first time t1 And put it into the following formula (4) to calculate the first measurement function of the distribution network at the first moment t1: and the first Jacobian matrix In formula (4), and The active power and reactive power are injected into the prediction nodes from the first node to the nth node among the n nodes of the distribution network within the real-time initial time t0; and The active power and reactive power are injected into the prediction nodes of the first branch to the mth branch of the distribution network within the real-time initial time t0 respectively; G ii+1 and B ii+1 are the conductance and susceptance between the i-th node and the i+1-th node among the n nodes of the distribution network, respectively; is the predicted voltage phase angle difference between the i-th node and the i+1-th node among the n nodes of the distribution network; and are respectively the predicted voltage amplitude at the head end and the predicted voltage amplitude at the tail end of the j-th branch among the m branches of the distribution network at the real-time initial time t0; G j and B j are the conductance and susceptance of the j-th branch among the m branches of the distribution network, respectively; is the predicted voltage phase angle difference between the head node and the tail node of the j-th branch among the m branches of the distribution network; Set the initial value of the first predicted state quantity of the distribution network at the first time t1 and the measurement noise covariance matrix; The real-time node voltage amplitude, real-time node injected active power, real-time node injected reactive power of n nodes of the distribution network within the first time t1 and the real-time active power and real-time reactive power of the branch head end of m branches are collected in real time to form a first measurement value The first prior state quantity First measurement function First Jacobian matrix The initial value of the first predicted state quantity and the first measurement value Substitute into the distribution network state estimation model to obtain the first predicted state quantity of the distribution network at the first time t1 As shown in the following formula (5), In formula (5), is the predicted node voltage amplitude from the first node to the nth node among the n nodes of the distribution network within the first time t1; is the predicted voltage phase angle from the first node to the nth node among the n nodes of the distribution network within the first time t1; The first predicted state quantity All predicted node voltage amplitudes and phase angles in are collected to form the first predicted node voltage phasor sequence The first predicted node voltage phasor sequence As the first state estimation result of the distribution network at the first time t1 Step 5: The first predicted state quantity As the second priori state quantity of the distribution network at the second time t2 And according to the principle of formula (4), the second measurement function of the distribution network at the second moment t2 is calculated as follows: and the second Jacobian matrix Setting the initial value of the second predicted state quantity of the distribution network at the second time t2 and the measurement noise covariance matrix; The real-time node voltage amplitude, real-time node injected active power, real-time node injected reactive power of n nodes of the distribution network at the second time t2 and the real-time active power and real-time reactive power of the branch head end of m branches are collected in real time to form a second measurement value The second prior state quantity Second measurement function Second Jacobian matrix The initial value of the second predicted state quantity and the second measurement value Substitute into the distribution network state estimation model to obtain the second predicted state quantity of the distribution network at the second time t2 As shown in the following formula (6), In formula (6), is the predicted node voltage amplitude from the first node to the nth node among the n nodes of the distribution network within the second time t2; is the predicted voltage phase angle from the first node to the nth node among the n nodes of the distribution network within the second time t2; The second predicted state quantity All predicted node voltage amplitudes and phase angles in are collected to form the second predicted node voltage phasor sequence The second predicted node voltage phasor sequence As the second state estimation result of the distribution network at the second time t2 Step 6: Repeat the principle of step 5 to obtain the distribution network from the third time t3 to the nth time t n The state estimation result of the distribution network is completed in the time period [t1,t n ] n state estimates within ].

2. The state estimation method according to claim 1, wherein: The step 4 also includes the following: The observability regions of n nodes in the distribution network are divided into node state observable regions Node ob and node status unobservable area Node un ; The first state estimation result Node in the node status observable area ob The predicted node voltage amplitudes and phase angles of the nodes in the are collected to form the first observable region voltage phasor prediction sequence U ob1 ; The first state estimation result The node status belongs to the unobservable area Node un The predicted node voltage amplitudes and phase angles of the nodes in the are collected to form the first unobservable region voltage phasor prediction sequence U un1 ; Establish the node status observable area Node un and node status unobservable area Node ob The state space mapping model between is shown in the following formula (7): In formula (7), U′ un is the voltage phasor correction prediction sequence of the unobservable area; H is the Jacobian matrix; U ob is the voltage phasor prediction sequence of the observable area; Y un is the admittance matrix of the unobservable region; Y L is a diagonal matrix whose diagonal elements are unobservable node admittances; Y u is the unobservable node admittance sequence; P un and Q un is the active power and reactive power of the unobservable node; U un Prediction sequence of voltage phasors for unobservable regions; The first observable area voltage phasor prediction sequence U ob1 and the first unobservable region voltage phasor prediction sequence U un1 Substitute into the state space mapping model to calculate the first unobservable area voltage phasor correction prediction sequence U' of the distribution network at the first time t1 un1 The first observable regional voltage phasor prediction sequence U ob1 and the first unobservable region voltage phasor correction prediction sequence U′ un1 Collected to form the first revised state estimation result of the distribution network at the first time t1 3. The state estimation method according to claim 1, wherein: The step 5 further includes the following: The second state estimation result Node in the node status observable area ob The predicted node voltage amplitudes and phase angles of the nodes in the are collected to form the second observable area voltage phasor prediction sequence U ob2 ; The second state estimation result The node status belongs to the unobservable area Node un The predicted node voltage amplitudes and phase angles of the nodes in the are collected to form the second unobservable region voltage phasor prediction sequence U un2 ; The second observable area voltage phasor prediction sequence U ob2 And the second unobservable region voltage phasor prediction sequence U un2 Substitute into the state space mapping model to calculate the second unobservable area voltage phasor correction prediction sequence U' of the distribution network at the second time t2 un2 The second observable regional voltage phasor prediction sequence U ob2 and the second unobservable region voltage phasor correction prediction sequence U′ un2 Collected to form the second revised state estimation result of the distribution network at the second time t2

Images