This application claims priority benefit of Chinese Application No. 200810223766.0 filed on Oct. 10, 2008.
The present invention relates to the field of power system analysis and calculation.
The state estimation, sometimes referred to as wave filtering, improves the accuracy of the data by using the redundancy of the real-time measurement system, eliminates the wrong information caused by the random disturbance automatically, estimates or forecasts the operation state of the system. The function of the state estimation is to convert the raw data with low accuracy, imperfection and bad data into processed integral and reasonable data with little error. The chief task of the state estimation is to construct a network model using the telemetry information and remote signaling information of the real-time measurement system according to switch state, and to make raw-detection on real-time measurement data to eliminate the bad data and supplement the deficient measuring points with pseudo measurement to ensure the redundancy, thus estimating the system state, calculating the power flow distribution, meanwhile analyzing the operation state of the system using calculated results, enhancing the observability of the whole network.
It has been more than 30 years since the traditional method for state estimation was established by Fred Charles Schweppe of MIT in 1970. Different state estimators have different estimation criterions according to different objective functions used during the process of resolving the state variables. Presently, the estimation criterions used in power system comprises WLS (weighted least square) criterion, non-quadratic criterion, WLAV (weighted least absolute value), LMS (least median of squares) and LTS (least trimmed squares). The most widely-used criterion is the WLS estimation criterion, which has such advantages as simple model, small computational amount, and excellent statistical properties including optimal un-bias property on estimating measurement values of the ideal normal distribution. However, the disadvantage is its bad capability of robust, that is to say, the estimation result is easily affected by bad data and is far from the true value. Non-quadratic criterion, WLAV, LMS and LTS estimator are all robust estimators having notably high capability of robust, but such disadvantages as large computational amount and requiring to subjectively determined weighted factors are important factors that hinder the development of the robust estimators.
The present invention provides a method for state estimation in power system, which provides an estimation criterion with good capability of robust that can reflect the real distributing condition of the measurement error, constructs an optimal mathematical model based on the above estimation criterion with improved correctness of estimation by adding the injection power of link nodes into constraint conditions as the equality constraints, gives a principle for setting parameters in objective functions and a method for setting an initial value when resolving the said optimization problem.
The objective of the present invention is to provide a method for state estimation based on real-time measurement in power system.
1. A method for state estimation based on the measured data of the data-acquisition system in power system is characterized by its implementation on computer as follows:
Block 1, acquiring the network parameters of the power system from the energy management system, the said network parameters comprising the branch number, head node number and end node number of the power transmission lines, the series resistance, the series reactance, the parallel conductance, the parallel susceptance, and the ratio and impedance of the transformer;
Block 2, acquiring the telemetry measurement value and the remote signaling measurement value from the network data server of Supervisory Control and Data Acquisition (SCADA), wherein,
The analog telemetry measurements comprise voltage amplitude, active power of the generator, reactive power of the generator, active power of the load, reactive power of the load, active power of the first end of the line, reactive power of the first end of the line, active power of the tail end of the line, reactive power of the tail end of the line and current amplitude of the line,
The remote signaling measurements comprise the switch state of the switches and the disconnecting links and the position of the transformer taps. The measurement is 1 when the switch or the disconnecting link is closed, and the value is 0 when the switch or the disconnecting link is open,
The data acquisition of the said SCADA system is implemented by polling mode with a cycle time of 5˜10 seconds, each of the measuring points includes at least the type, the point number and the measurement value;
Block 3, constructing a computational model;
Block 3.1, defining the ports of the switches, the power transmission lines, the transformers, the generators, the capacitors and the loads as nodes, traversing these nodes, merging the nodes connected together by closed switches or disconnecting links in the network into one node used for calculation according to the switching values, all the following nodes refer to the nodes used for calculation,
Block 3.2, distributing the voltage amplitude measurements, the generator power measurements and the load power measurements acquired from block 2 to their corresponding nodes according to the ways as follows:
Averaging to get the voltage amplitude measurement value if the said nodes include a plurality of voltage amplitude measurement values,
Adopting generator power from which the load power is taken away to be the injection power of the node, only if the power measurement happened on all the generators and loads on the said node;
Block 4, taking the voltage amplitude and voltage phase angle of the node as state variable x, and taking one of the nodes as the reference node of the voltage phase angle, then resolving the following equations to get an optimal solution “x*” of the said state variable:
Block 4.1, defining the nodes without a generator nor a load as link nodes, the said link nodes form a set A while the generator nodes form a set E, defining Pkmax and Qkmax to indicate the maximum active power output and the maximum reactive output of the generator node k respectively,
Block 4.2, resolving the optimal resolution x* by adopting following optimization model:
Wherein, A is a set consisting of link nodes, E is a set consisting of generator nodes,
for the voltage measuring points: αi=0.02×|zibase|,
for the active power measuring points: αi=0.02×|zibase|,
for the reactive power measuring points: αi=0.03×zibase|,
The value of zibase is related to the i-th measurement and is predetermined corresponding to different voltage levels,
βi is the parameter related to the said αi,
βi=c/αi, c is a constant, c=5˜10;
Block 5, calculating the estimated value by using the formula z*=h(x*) according to the optimal solution x* derived from block 4,
The voltage amplitude of node i is: h(x*)=ν*i,
The active injection power of node i is:
The reactive injection power of node i is
ν*i, ν*j, θ*i, θ*j are all the optimized solutions of the state variable which can be represented by x*;
Block 6, determining whether each of the estimated values of the measurement is qualified or not according to the following formula:
If |zi−zi*|≦αi, the estimated value is qualified.
2. The method for state estimation based on the measured data of the data-acquisition system in power system of claim 1, characterized in that, the said zibase under the line power measurement values as follows: values 1083 MVA under the voltage level of 500 kv, values 686 MVA under the voltage level of 330 kv, values 305 MVA under the voltage level of 220 kv values 114 MVA under the voltage level of 110 kv, values 69.7 MVA under the voltage level of 66 kv.
3. The method for state estimation based on the measured data of the data-acquisition system in power system of claim 1, characterized in that, the said zibase under the voltage amplitude measurement values as follows: values 600 kv under the voltage level of 500 kv, values 396 kv under the voltage level of 330 kv, values 264 kv under the voltage level of 220 kv, values 132 kv under the voltage level of 110 kv, values 79.2 kv under the voltage level of 66 kv.
4. The method for state estimation based on the measured data of the data-acquisition system in power system of claim 1, characterized in that, the said zibase under the generator power measurement values apparent power of the generator.
The proposed method for state estimation can be applied to EMS of regional power and provincial power network, and can also be applied to the monitoring, analysis and control system of power system based on real-time measured data.
SCADA, the Supervisory Control and Data Acquisition system is a real-time data source for power system automation which provides real-time measured data of power system. The SCADA measurement includes the telemetry measurement indicative of measurement of analog quantity and the remote signaling measurement indicative of measurement of discrete quantity. The telemetry measurement equation in power system can be represented as follows:
z=h(x)+ν (1)
Wherein, z is the vector of the measurement value; h(x) is the vector of the measurement function; ν is the vector of the measurement error; assuming the number of the total measurements as m, then all the above vectors have m dimensions; x is a state variable, assuming the number of the nodes in the system as n, adopting the voltage amplitude and the voltage phase angle of the nodes as the state variables, and taking one of the nodes as the reference node of the voltage phase angle, then x has 2n−1 dimensions, equation (2) to equation (12) are measurement functions for some common measurements respectively:
The injection power Pi, Qi of node i can be represented as follows:
The power flow of node i side on line i-j can be represented as follows:
Pij=νi2g−νiνjg cos θij−νiνjb sin θij (4)
Qij=−νi2(b+yc)−νiνjg sin θij+νiνjb cos θij (5)
The power flow of node j side on line i-j can be represented as follows:
Pji=νj2g−νiνjg cos θij+νiνjb sin θij (6)
Qij=−νj2(b+yc)+νiνjg sin θij+νiνjb cos θij (7)
The power flow of node i side on transformer branch i-j can be represented as follows:
The power flow of node j side on transformer branch i-j can be represented as follows:
The voltage amplitude measurement of node i can be represented as follows:
Ui=νi (12)
νi, νj in the measurement functions above are voltage amplitudes of node i and node j respectively, θij is the voltage phase angle difference between node i and node j, Gij, Bij are admittance matrix elements, g, b, yc are parameters in π-type model of the line, k is the non-standard voltage ratio of the transformer, bT is the standard measured susceptance of the transformer, the π-type equivalent circuit diagram of the line and the transformer is shown in
The nodes without neither a generator nor a load are referred to as link nodes, assuming a set A consisting of link nodes, a set E consisting of generator nodes, Pkmax and Qkmax are used to denote the maximum active power output and the maximum reactive power output of the i-th generator, respectively, adopting the optimization data model as follows:
αi, βi in equation (13) are parameters related to the i-th measurement which value according to the following principles
1. αi values according to principles as follows
measuring points for voltage measurement: αi=0.02×|zibase|
measuring points for active power measurement: αi=0.02×|zibase|
measuring points for reactive power measurement: αi=0.03×|zibase|
the value of zibase relates to the i-th measurement and values according to following principles:
a) zibase for the line power measurement, values 1083 MVA under the voltage level of 500 kv, values 686 MV under the voltage level of 330 kv, values 305 MVA under the voltage level of 220 kv, values 114 MVA under the voltage level of 110 kv, values 69.7 MVA under the voltage level of 66 kv;
b) zibase for the voltage amplitude measurement, values 600 kv under the voltage level of 500 kv, values 396 kv under the voltage level of 330 kv, values 264 kv under the voltage level of 220 kv, values 132 kv under the voltage level of 110 kv, values 79.2 kv under the voltage level of 66 kv;
c) zibase for the generator power measurement values generator apparent power.
2. βi values according to the principles as follows:
βi=c/αi, c is a constant which values ranging from 5 to 10 in the present invention.
When the deviation between the estimated value and the measurement value in the i-th measurement meets the formula that |zi−hi(x)|≦αi, the estimated value of the measuring point can be considered as qualified according to the SGCC (State Grid Corporation of China) standard used for determining whether the estimated value of the measuring point is qualified.
Assuming the optimum solution as x* and calculating the estimated value of the measurement using the formula z*=h(x*), then the whole computation is finished.
The present invention can be applied in the real-time control system of practical power system named AEMS, the program flowchart of which is illustrated in
The detailed description of the steps is set forth as follows by taking the 4-node system shown in
Step 1: Initialization
Acquiring the network parameters of the power system from existing energy management system, the said network parameters comprise the series resistance, series reactance, parallel conductance and parallel susceptance of the transmission line, the ratio and impedance of the transformer.
The line parameters of the 4-node system are shown in table 1:
Step 2: Acquiring the SCADA Measurements
The measurements comprise the telemetry measurement and the remote signaling measurement. The telemetry measurements mainly comprise the voltage amplitude measurement, the generator active power measurement, the generator reactive power measurement, the load active power measurement, the load reactive power measurement, the reactive power measurement of the header point of the line, the active power measurement of the end point of the line, the reactive power measurement of the end point of the line, the current amplitude measurement of the line. The remote signaling measurements mainly comprise the measurement of the switch state of the switches and the disconnecting links and the position of the transformer taps. Each of the measuring points at least comprises the type, the serial number and the measurement value. The above-mentioned measurement value can be given by the file under offline condition, and derived from the SCADA system by API or TCP/IP communication protocol provided by SCADA system under online application condition. The SCADA system acquires data by polling mode with a usual cycle of 5˜10 seconds.
Assuming the measurement data of the 4-node system at a particular time is shown in table 2, wherein, when the remote signaling measurement value is 1, it indicates that the switch is closed; when the remote signaling measurement value is 0, it indicates that the switch is open. The remote signaling measurement value is represented by named value, e.g. the units of the active power, the reactive power, the voltage amplitude and the angle are MW, MVAR, KV and “∘” respectively.
Step 3: Forming a Computational Model
The ports of the switches, transmission lines, transformers, generators, capacitors and loads are referred to as nodes. Traversing all these nodes, merging the ones that are connected together by closed switches or disconnecting links in power network to be one node used for calculation according to the switching value, the nodes mentioned as follows all indicate the nodes used for calculation.
Redistributing the voltage amplitude measurement value, generator power measurement value and load power measurement value acquired from step 2 onto the nodes specifically by methods as follows: 1) averaging to get the voltage amplitude measurement values if the nodes have a plurality of voltage amplitude measurement values, 2) adopting the generator power from which the load power is taken away to be the injection power of the node, only if the power measurement happened on all the generators and loads on the node, then entering into the next step.
The bus lines 1-1 and 1-2 in the 4-node system are connected by switch K2, which emerge into a bus under the condition that K2 is closed, the said bus has a voltage amplitude measurement that is the average value of the voltage amplitude value on 1-1 and 1-2. The calculation model is shown in
Step 4: Resolving the State Variables
Assuming a set A consisting of connection nodes, a set E consisting of generator nodes. Pkmax and Qkmax indicate the maximum active power output and maximum reactive power output respectively. Adopting an optimization data model described as follows:
The state variables derived from estimation by WLS estimator can be considered as the initial values under the condition of offline application, and the state variables derived from the previous estimation can be considered as the initial values under the condition of online application. Adopting the resolution method of the nonlinear optimization problem to resolve the value and then assuming the optimal solution as x*.
Node 1 is considered as the referenced node of the voltage angle in the 4-node system, the vector z of the measurement value and parameters such as αi, βi are shown in table 4. The results of the calculation for the state variables are shown in diagram 5.
Step 5: Calculating the Estimated Values of all the Measurements
Resolve the estimated value of the measurement value by using the formula z*=h(x*) and preserve the calculation results shown in table 6, wherein the said x* is derived from the previous step. Hold on for a few seconds and return to step 2 under the condition of online real-time calculation; the whole calculation is finished under the condition of offline calculation.
Number | Date | Country | Kind |
---|---|---|---|
2008 1 0223766 | Oct 2008 | CN | national |
Number | Name | Date | Kind |
---|---|---|---|
3903402 | Petit et al. | Sep 1975 | A |
20090228154 | Trias | Sep 2009 | A1 |
Number | Date | Country | |
---|---|---|---|
20100094574 A1 | Apr 2010 | US |