RU2809081C1 - Method for monitoring low-voltage power grid based on accems - Google Patents

Abstract

FIELD: electrical engineering.

SUBSTANCE: invention relates to the field of electrical engineering and can be used for operational monitoring of 0.4 kV power grids. The technical result is an expansion of the functionality of the current automated commercial electricity metering system (ACCEMS) with the ultimate goal of reducing electricity losses, in particular, identifying and localizing points of uncontrolled consumption (UC) of electricity on the circuit, identifying consumer phases, accurate control of technically justified losses taking into account the temperature dependence of the resistance of wires, monitoring currents and voltages of all network elements. It is achieved by creating a network model based on a priori information about its circuit, lengths and parameters of wires. On entry into the network the average voltage and ambient temperature are measured at the metering interval synchronized with the system time of the ACCEMS. Based on the readings of the ACCEMS, the averages are calculated at the interval of accounting for consumer power and at the network input. The actual power imbalance is calculated as the difference between the input power and the sum of the consumer capacities. According to the readings of power and voltage at the network input, the current of the head section is calculated. The consumer currents are determined numerically and technically justified losses in the network are calculated, taking into account the dependence of the resistance of the wires on the ambient temperature and the flowing current. Calculate the discrepancy of the power imbalance of the network as the difference in the actual unbalance and technically justified losses. The detection of the fact of UC is fixed on the basis of the current imbalance. In order to localize UC points and identify consumer phases, measurements and calculations are repeated at multiple time intervals. The current imbalance is alternately added to each node of the network model and the corresponding power imbalance residuals are calculated. The desired UC point will correspond to a network node for which the power imbalance discrepancy in the entire statistical sample is minimal. The identification of consumer phases is performed on the basis of calculating the correlation coefficients of time series of phase capacities at the input and at consumers. After eliminating the UC and identifying the phases of consumers, current control of technically justified losses, average currents, voltages and temperatures of wires is carried out.

EFFECT: expansion of the functionality of the current automated commercial electricity metering system (ACCEMS).

1 cl, 4 dwg

Description

The invention relates to the field of electrical engineering and can be used for operational monitoring and detection of uncontrolled consumption (UC) of electricity in 0.4 kV electrical networks. The inventive method makes it possible to implement the following functions of operational network monitoring:

- Detect the fact and localize the place of uncontrolled consumption on the diagram;

- Identify consumer phases;

- Determine technically justified losses for current weather conditions and network load;

- Monitor the ratio of temperatures and wire currents to their long-term permissible values in all sections of the network;

- Monitor the average voltages and currents of all network elements over the metering interval.

The main purpose of the invention is to build a fully functional and effective network monitoring system based on the current automated control system, without the need to significantly expand the composition of technical means (use of additional devices). Wherein:

- network monitoring is carried out at the rate at which ASKUE data is updated, for example, every hour;

- from AMR measurements, only data on cumulative totals of electricity with time stamps are used (without the need to organize a complex process of synchronous measurements of currents and voltages).

1. Prototypes (state of the art)

There is a known method for identifying and monitoring non-technical losses in 0.4 kV distribution networks (Patent RU 2 757 655 C1; G01R 11/24, H02J 13/00; 2021.08), which consists in measuring the current and voltage vectors with the main metering device and metering devices for all subscribers. Based on the measurements performed, complex values of currents in the phases of the main sections of the distribution network that supply its last subscribers are calculated. In this case, the fact and magnitude of uncontrolled electricity consumption is determined by comparing the calculated currents in the phases of the main sections of the distribution network with the actual complex current values measured by the metering devices of the last subscribers of the controlled section of the network. To monitor non-technical losses of electricity, a prompt calculation of the complex resistance of inter-subscriber sections of the network is organized in the absence of currents of uncontrolled electricity consumption based on data from two different operating modes. Uncontrolled electricity consumption is detected by analyzing the complex resistances of inter-subscriber network sections, which will differ from the actual ones found previously in the absence of uncontrolled consumption.

The disadvantages of this method are:

- The problem of time synchronization over communication channels with packet data transfer protocols (intermediate accumulation and uncontrolled delay);

- The need for preliminary identification of subscriber phases, which is a priori unknown;

- The problem of obtaining measured current and voltage vectors in the case when standard digital electricity meters with a communication interface are used as subscriber devices;

- The need for constant monitoring of the network, since during periods of absence of uncontrolled consumption, identifying its place in the network in this way is impossible;

- The problem of high traffic when using paid or low-speed channels.

There is a known method for analyzing metering points in order to identify unaccounted for volumes and search for electricity losses, based on expert algorithmic assessment of load profiles using artificial intelligence (AI).

This approach does not yet provide acceptable reliability of target designation (TD) to the NP point (less than 30%), and also requires significant computational costs and a long time for training the neural network.

The main task to be solved by the claimed invention is to eliminate the indicated disadvantages of the prototype methods.

2. Technical support of the method

Standard digital electricity meters with a communication interface are used as subscriber monitoring devices, without the need to expand their functionality and use other measuring instruments (MI).

Using any available channels and standard communication protocols, the meters transmit information to the upper level of ASKUE, which includes a system for providing a unified system time (SOEV).

A balance meter is installed at the network input and synchronized with the ASKUE system time.

Voltage measurement is necessary only at one network node: at its input. This operation can be performed either by a separate meter or by a balance meter with the specified function. In this case, it is not the instantaneous or effective voltage value that is measured, but its average value over the energy metering interval of the ASKUE. If it is not possible to measure the average voltage, it is replaced by the nominal value. In this case, the estimates of all parameters turn out to be somewhat biased, but this does not lead to a failure of the method.

The temperature meter can be a corresponding information system, for example, Gismeteo, or a separate digital thermometer. For cable lines laid in the ground, the ambient temperature is taken from tables of average monthly soil temperatures for the corresponding geographic coordinates of the area and date.

In the dispatch center, a computer with application software is installed and connected using standard communication protocols: with the upper level of the ASKUE, balance meter, medium voltage and ambient temperature meters. All specified technical measuring instruments are synchronized with the AMR system time. De-synchronization of the meters' own time is allowed, which in practice can reach several minutes.

Unlike AI-based methods, the computing power requirements are relatively small. A single-board computer, workstation (PC), PLC can cope with the simultaneous monitoring of several dozen controlled 0.4 kV networks. The software can also be installed directly on the ASKUE server.

The block diagram of technical means is presented in Fig. 1.

3. Mathematical and software method

Composition of the software package

The composition of the software package is shown in Fig. 2.

Network emulator

A 3-phase network emulator is based on the strict laws of electrical engineering, taking into account the reactivity of the resistance of wires and loads (calculations in the field of complex numbers) and the dependence of the active resistance of the lines on the flowing current and ambient temperature. Module purpose:

- Modeling of modes and calculation of network parameters entered into the electrical diagram editor;

- Testing and research of modules working with real data;

- Generation of missing or replacement of invalid data.

Calculation module

For the number of loads greater than two, it is impossible to write expressions for currents in analytical form as an explicit function of load power, voltage at the network input and wire resistance. Therefore, current calculations are performed by selecting them using numerical methods. The minimized objective function is composed by any known calculation method from the theory of electrical circuits: the method of nodal potentials, loop currents, etc.

Taking into account the dependence of wire resistance on weather conditions and flowing current

The wire resistance under operating conditions, relative to the average value of 25°, can vary within the following limits:

Where:

- temperature coefficient of resistance (0.0043 1/°C for aluminum); - operating temperature range of the wire (-40°…+90°).

Ignoring the dependence of wire resistance on weather conditions and flowing current can lead to unacceptable errors in both monitoring parameters and errors in detecting the fact of uncontrolled consumption and targeting its coordinates in the network.

It is not possible to write down an analytical formula for the explicit dependence of the wire temperature on the flowing current.

A simplified technique is based on the linearization of the nonlinear and implicit dependence of the wire temperature on the flowing current and ambient temperature (Fig. 3)

As calculations show, the maximum deviation ΔT° of linearization does not exceed 7°C, which leads to an acceptable additional error in calculating the resistance of an aluminum wire - no more than 2.8% in the range of operating conditions.

Now, knowing the current, ambient temperature, long-term permissible values of temperature and current of the wire, using an explicit straight line equation, you can easily and quite accurately calculate the temperature of the wire and its current resistance.

Uncontrolled consumption detection and localization module

The main problem of instrumental methods for detecting IR is the very low resistance of the wires, which leads to subtle voltage drops at the internode sections of the network, which turn out to be less than the voltage measurement error. This problem is solved with the help of redundancy of measuring information - by moving to statistical analysis of the measuring array.

Another problem is the unknown distribution of consumers across network phases. In practice, the vast majority of network passports and their support circuits do not contain this information. Therefore, first a completely symmetrical network model is adopted, in which the neutral current is zero. In this case, a transition is made to a single-phase equivalent (single-line diagram) of the network.

Let us introduce the following notation for each time interval t=1,…,T :

- average power at the network input (n=0) and legal consumers ( ); - average power of NP; - average voltage at the network input; - current of the head section (n=0); - numerical assessment of consumer currents ( ); - current in the network section ; calculated according to Kirchhoff's 1-law based on known currents of legal loads and network diagram; - resistance of the network section taking into account the wire temperature , depending on the ambient temperature and flowing current; N - the number of legal consumers on the network. L - number of network sections (main, branches and taps)

The main controlled network parameters are calculated using the formulas:

Current imbalance: (1) Power imbalance: (2) Technically justified (legal) losses: (3) Power imbalance discrepancy: (4)

The power balance equation in the network has the form:

(5)

Where - NP current in the network section .

The second term on the right side of (5) determines the amount of additional losses due to accidents in sections of the network. From (4) and (5) it is easy to see that at zero current and power of the NP, the unbalance discrepancy is also equal to zero. The third term on the right side of (5) is the value of the NP power, which can be calculated based on the Kirchhoff and Joule-Lenz laws for each network node to which a current imbalance is added. The results of calculating the unbalance using formulas (4) and (5) for the simplest circuit are presented in Fig. 4.

Thus, (subject to the difference in line resistance for all loads), the set of known values - legal currents, NP current and unbalance mismatch - corresponds to one and only one point on the diagram, when adding a current unbalance to which, a zero power unbalance mismatch is ensured.

However, it is impossible to reliably calculate it from one reading, since due to: measurement errors, inaccuracy in specifying wire lengths, inequality of actual wire parameters to their rated values, phase asymmetry, the presence of NPs, the current estimates will be shifted. It is possible to obtain a reliable (in a probabilistic sense) control center based on multiple measurements and subsequent statistical processing of redundant measurement information.

The optimal sampling depth T can be considered an array of hourly intervals over a week, which forms a repeating pattern of load profiles.

Thus, the target designation (TS) is such a point on the diagram for which:

(6)

Identification of consumer phases

Optimal distribution of loads across the phases of the supply feeder is a necessary condition for improving the symmetry of the 3-phase network and reducing losses in the neutral. A strict, phase-by-phase calculation of losses and operating parameters of the network requires precise knowledge of which phase of the power line a specific subscriber is connected to. A priori information about the correspondence of the inputs (L1, L2, L3) of subscriber 3-phase and single-phase meters to a specific phase (A, B, C) of the 0.4 kV distribution network is, as a rule, absent.

To solve this problem, the results of multiple measurements are also used - numerical series of load power profiles and at the network input.

The phase of the network to which the load is connected can be preliminarily determined by the highest correlation coefficient (Pearson) between the numerical series of phase readings of the load meters and the balance meter. At the same time, for 3-phase loads, the most loaded phase is selected first, and the remaining two are determined by the known phase rotation, which is available for any modern meter. Next, checksums are checked and, if necessary, phases are rearranged until the search stop condition is met:

(7)

Where:

- Total power of the network phase; ( ); - Consumer power at input L, in the current arrangement assigned to phase Ф ( ).

It is important to note that the accuracy of load phase identification significantly depends on the presence of uncontrolled consumption in the network, so it is performed after eliminating illegal network loads.

4. Checking the functionality of the method

The inventive method was tested both on synthetic data and in real 0.4 kV power networks. The method showed high reliability of the CC-NP and sensitivity: in one of the real networks an unaccounted for street lamp was detected and an accurate target indication of its location was obtained.

Claims (1)

A method for monitoring a 0.4 kV electrical network based on an automated electricity metering control system (ASCAE), which consists in creating a digital model of the network using a priori information and measuring the increments of electricity at the input to the network and at consumers at identical time intervals; characterized in that in order to detect and target the location of uncontrolled consumption in the network, identify load phases and increase the accuracy of calculating technically justified losses at identical time intervals: measure the average voltage at the input to the network, measure the ambient temperature; calculate the average power at the input to the network and at consumers, select the average values of consumer currents using numerical methods, compare the sum of these currents with the value calculated as the ratio of the average power to the average voltage at the input to the network, in the case of a significant difference in the imbalance of currents, record the fact of uncontrolled consumption; calculate the average power of technically justified losses in the network, taking into account the network model and the dependence of wire resistance on the flowing current and ambient temperature, calculate the difference between the power input into the network and the sum of consumer powers (average power imbalance), calculate the difference between the actual power imbalance and the calculated losses (imbalance discrepancy); These operations are repeated for a sufficient number of time intervals, the current unbalance is alternately added to each node of the network model and the corresponding power imbalance residuals are calculated, mathematical processing of the temporary array of measurement information obtained in previous operations is performed, as a result of which target designation is formed for points in the network diagram with unaccounted consumption; eliminate unaccounted consumption; repeat the above operations and, in the absence of uncontrolled consumption, perform mathematical processing of a temporary array of measurement information, as a result of which load phases are identified; calculate technically justified losses in all network elements and its operating parameters for all time intervals.