KR20120137613A - Method for assume condition of power distribution system - Google Patents

Abstract

PURPOSE: A method for estimating the state of a power distribution system is provided to accurately check current problems by checking the current state of the power distribution system based on a mathematical analysis result after performing an accurate power flow calculation algorithm. CONSTITUTION: Injection power of a substation is set. A voltage of a load bus line is set as 1.0 pu. Unknown load power is set as load capacity. An active power assignment coefficient and reactive power assignment coefficient for load distribution are set as 1. Power flow calculation is performed after load distribution by state estimation. [Reference numerals] (AA) Start; (BB,KK) Setting unknown load power; (CC) Calculating bus power; (DD) Inputting data; (EE,OO) Calculating a loss function; (FF) Forming Ybus of a system; (GG) Initial stage; (HH) Setting an initial value of the system; (II) Calculating a load power assigned coefficient; (JJ) Calculating a voltage V_NL by a voltage source; (LL) Distributing load; (MM) Repeated calculation step; (NN) Calculating power flow; (PP) Calculating an injection current of a load bus; (QQ) Loss function convergence?; (RR) Calculating a voltage by a current source; (SS) Outputting a result; (TT) End; (UU) Synthesizing a voltage result; (VV) Power flow calculation convergence?

Description

Method for assume condition of power distribution system

The present invention relates to a method for estimating a state of a power distribution system. More specifically, the load of an unknown bus is calculated by reflecting the drawing value of a substation and known values by measurement of some buses, and calculating a tidal current of the power distribution system using the same. The present invention relates to a method for estimating a state of a power distribution system in which load distribution can be reasonably calculated and corrected by repeating the process.

In recent years, the power system has become larger in size and the linkage of the tracks has been extensive. Due to this size and complexity, the cost of investment in power equipment is increasing, and the amount of power consumption is also increasing, and the overall cost including fuel costs for power equipment is increasing year after year.

84% of Korea's greenhouse gas emissions are emitted from the energy sector, and electric power accounts for 34% of them. Moreover, global electricity demand is expected to increase by about 70% in 2020 and twice in 2030 compared to 2006, and the necessity of intelligent and convergence technologies for power grid for efficient management is increasing day by day.

KEPCO's transmission and distribution loss rate is 16,770GWh, which is 4.07% of 2009's 411,631GWh. Since the loss rate of the distribution system accounts for more than 50% of the transmission and distribution losses, the distribution loss amount is about 8,400 GWh. Therefore, when the system optimization of the distribution line (D / L, feeder) is promoted, the loss reduction can solve about 10% of the distribution loss rate. Then, as of 2009, the loss reduction achieved through power distribution system optimization is equivalent to 840GWh, and if it is converted into about 20 won, which is the cost of coal power generation per kW, it is estimated that the amount of savings of about 140 billion won can be saved annually. .

The analysis technology of the distribution system, the core technology for optimizing the power system, and the most basic of them, power flow calculation is the cornerstone for the power grid intelligence.

The tide calculation is to perform a mathematical analysis of the power system in a steady state, usually interpreted using a software tool. The result of the tide analysis is called base-case, which includes voltage information of all nodes, current information of the line, and grid loss information.

This information is used not only for the optimal operation of the power system, but also for the planning of future grid growth. Since the tidal current calculation program for power transmission system became practical in the late 1950s, various analysis methods such as Newton-Raphson method, Gauss-Seidel method, and Fast Decoupled Power Flow method have been developed.

The analysis method is based on the assumption that the lines of the transmission system are evenly (transposed 3-phase lines) and that the a-b-c phases are in equilibrium in steady state. Therefore, in the transmission system, the system is treated as a single phase and modeled, so that the transmission system can be analyzed relatively simply.

However, in actual distribution systems, lines are rarely arranged in series, and most of them are composed of three-phase, two-phase and one-phase lines or an unbalanced system in which loads are mixed. Therefore, three-phase modeling is essential for correct analysis of the distribution system.

On the other hand, in addition to considering the unbalanced power distribution system, unlike the power transmission system, a situation in which accurate load information is not known because power values of all loads are not given in actual operation. This is because the load power is measured only where the measurement equipment is included, such as substations or division switches.

That is, the load is not known where the manual switch without the measuring device is located or far away from the division switch. In order to grasp the status of the distribution system, a tidal current calculation should be performed. In the distribution system, there is little information on the active power and reactive power of the load.

Thus, unbalanced load sharing of the power distribution line is caused, which increases the loss of the power distribution system, and sometimes involves current or voltage problems. Therefore, in order to overcome this problem of the distribution system, it is necessary to apply the Implicit Zbus method suitable for the distribution analysis in order to calculate the unknown load and to study the load distribution analysis algorithm for the distribution system.

Accordingly, an object of the present invention is to calculate the load of the unknown bus by reflecting the withdrawal value of the substation and known values by the measurement of some buses, since there is little information on the active power and reactive power in the actual operation of the distribution system. It is to provide a state estimation method of a power distribution system in which load distribution can be reasonably calculated and corrected by repeatedly performing a bird flow calculation of a power distribution system.

)을 계산하는 초기화 단계; 및 상태추정에 의해 부하 배분을 수행한 후 조류 계산을 수행하고, 편차를 확인하여 수렴 여부를 판정하는 반복 계산 단계로 이루어진 것을 특징으로 한다.The state estimation method of the power distribution system according to the present invention for achieving the above object of the present invention, after configuring the Ybus matrix of the power distribution system and setting the system initial value,

An initializing step of calculating; And it is characterized in that it consists of an iterative calculation step of determining the convergence by performing a bird flow calculation after performing the load distribution by the state estimation, checking the deviation.

In this case, setting the initial value of the system, a) setting the injection power of the substation as the total supply power, b) setting the voltage of the load busbar to 1.0pu as the known voltage of the substation, c A) setting an unknown load power to a load capacity; and d) setting an active power allocation coefficient and a reactive power allocation coefficient to 1 for load distribution.

)을 설정하는 단계와, b) 모선 전력을 계산하는 단계와, c) 손실함수를 계산하는 단계와, d) 유효전력 배정계수와 무효전력 배정계수를 계산하는 단계와, e) 상기 미지부하의 전력(

)을 배분하는 단계로 이루어진 것을 특징으로 한다.The load distribution may include a) power of an unknown load (

B) calculating bus power, c) calculating loss function, d) calculating active power allocation factor and reactive power allocation factor, and e) the unknown load. power(

) Is characterized by consisting of the step of distributing.

)를 계산하는 단계와, b) 전류원에 의한 전압(

)을 계산하는 단계와, c) 전압 결과를 합성(

)하는 단계와, d) 전압값의 이전 값과 갱신된 값의 편차가 허용 편차 내에 있으면 조류 계산 단계를 중단하고, 상기 허용 편차 내에 없으면 상기 조류 계산 단계를 반복 계산하는 단계로 이루어진 것을 특징으로 한다.In addition, performing the tidal current calculation, a) the injection current of the load bus (

) And b) the voltage by the current source (

) And c) synthesize the voltage results (

And d) stopping the algal flow calculating step if the deviation between the previous value and the updated value of the voltage value is within the allowable deviation, and repeatedly calculating the algal calculating step if it is not within the allowable deviation. .

On the other hand, the step of determining the convergence by checking the deviation, after calculating the loss function with the updated voltage (V), if the deviation between the previous value of the loss function and the update value of the loss function is within the allowable deviation, If it is not within the allowable deviation, it is characterized in that it is repeated until the deviation of the loss function reaches the allowable level.

According to the method of estimating the state of the power distribution system as described above, since there is little information on the active power and the reactive power in the actual operation of the distribution system, the load of the unknown bus is reflected by reflecting the drawing value of the substation and the known values by measuring some buses. By calculating the and repeating the process of performing the algae calculation of the distribution system using this there is an effect that the load distribution can be reasonably calculated and corrected.

In particular, the present invention can satisfy the power supply and demand conditions by repeatedly calculating until the deviation of the loss function and the individual voltage reaches the allowable level, the load power allocation coefficient for it can be obtained and the distribution of the unknown load can be reasonably performed. In addition, the loss of the distribution system is reduced, and after performing an accurate tidal calculation algorithm reflecting the load distribution, the current status of the distribution system is accurately identified based on the mathematical analysis results, and the problems and violations of the current state are identified. It also has the effect of eliminating the findings and operating reasonably.

1 is a flowchart of a state estimation method of a power distribution system according to an embodiment of the present invention.

While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. It should be understood, however, that the invention is not intended to be limited to the particular embodiments, but includes all modifications, equivalents, and alternatives falling within the spirit and scope of the invention. Like reference numerals are used for like elements in describing each drawing.

The terminology used herein is for the purpose of describing particular example embodiments only and is not intended to be limiting of the present invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this application, the terms "comprise" or "have" are intended to indicate that there is a feature, number, step, operation, component, part, or combination thereof described in the specification, and one or more other features. It is to be understood that the present invention does not exclude the possibility of the presence or the addition of numbers, steps, operations, components, components, or a combination thereof.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art. Terms such as those defined in the commonly used dictionaries should be construed as having meanings consistent with the meanings in the context of the related art and shall not be construed in ideal or excessively formal meanings unless expressly defined in this application. Do not.

The present invention can be embodied as computer readable codes on a computer readable recording medium. The computer-readable recording medium includes all kinds of recording devices in which data that can be read by a computer system is stored. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, and the like, and are also implemented in the form of a carrier wave (for example, transmission over the Internet). It also includes. The computer readable recording medium can also be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion.

Hereinafter, with reference to the accompanying drawings, it will be described in detail a preferred embodiment of the present invention.

In order to find out the status of the distribution system, a tidal current calculation should be performed. In the distribution system, there is little information on the active power and reactive power of the load. However, it is possible to estimate the load of the unknown bus by reflecting the subtracted value of the substation and the known values by measurement of some buses.

In the distribution system, the following information is used as a known value.

,

), 유효 및 무효 전력(P, Q)-변전소 모선은 발전 모선으로 취급Voltage magnitude and phase angle of substation (

,

), Active and reactive power (P, Q) substation busbars are treated as generating busbars

˚ Combined capacity of load stage transformer (kVA)

) - 역률이 주어지지 않으면 0.9 지상을 가정˚ Power factor of all load buses

)-0.9 ground if no power factor is given

In this way, the state estimation algorithm can perform the distribution system algae calculation using the values known by the measurement and the load calculation result. By repeating this procedure, the load distribution can be reasonably calculated and corrected.

)를 이용하여 계산한다.First, the load distribution algorithm for load distribution is a loss function (1) which can be expressed by Equation 1 below.

Calculate using

은 시스템의 총손실,

는 변전소의 총 공급전력,

는 조상기의 총 전력,

는 선로의 병렬 커패시턴스에 의한 총 충전 전력,

는 기지 부하의 총 전력,

는 미지 부하의 총 전력이다.only,

Is the total loss of the system,

Is the total supply power of the substation,

Is the total power of the ancestors,

Is the total charging power due to the parallel capacitance of the line,

Is the total power of the base load,

Is the total power of the unknown load.

Since it is difficult to distribute the load accordingly because the exact loss of the power distribution system is not known, the method of distributing the load so that the loss of the system is constant by repeated calculations is applied. Allow distribution to be corrected.

1 is a flowchart of a state estimation method of a power distribution system according to an embodiment of the present invention.

)을 수학식 2를 이용해 계산한다. Referring to FIG. 1, the load distribution algorithm may be largely divided into an initialization step and an iteration calculation step. First, in the initialization step, the Ybus matrix of the distribution system is configured, and after setting the system initial value,

) Is calculated using Equation 2.

At this time, in the step of setting the system initial value, the injection power of the substation is set as Equation 2 as the total supply power.

)으로 부하 모선의 전압은 1.0pu로 설정하고, 미지 부하전력은 수학식 3과 같이 부하 용량으로 설정하며, 부하 배분을 위한 유효전력 배경계수(

) 및 무효전력 배정계수(

)를 1로 설정한다. The voltage at the substation is known

), The voltage of the load bus is set to 1.0pu, the unknown load power is set to the load capacity as shown in Equation 3, and the effective power background coefficient (

) And reactive power allocation factor (

) To 1

Next, the iterative calculation step is composed of a large step of performing load distribution, performing a tidal current calculation, and determining the convergence by checking the deviation.

In the performing of the load distribution, the power of the unknown load is set as in Equation 5.

Then, after calculating the bus power as shown in Equation 6, the loss function represented by Equation 7 is calculated.

In this case, the loss function is calculated by calculating the voltage value by converting the load power and the power of the ancestor into an equivalent injection current source by setting the substation as a voltage source, and thus, Equations 2 and 8 when the voltage source and the current source are activated respectively are synthesized. Equation 9 is obtained.

As a result, since the product of the Ybus matrix and the predetermined V becomes the injection current I, the injection current S becomes as shown in Equation 10. When the injection powers of all buses are calculated and the injection powers of all buses are summed together, Equation 7 Total loss.

After calculating the loss function, the load power allocation factor is calculated.

In Equation 1, the total power of the load may be divided into active power and reactive power as shown in Equations 11 and 12.

은 부하의 총 유효 및 무효 전력이고,

는 i번째 모선의 유효 및 무효 전력이며,

는 값을 알고 있는 k번째 모선의 유효 및 무효 전력이고,

는 값을 알지 못하는 u번째 모선의 유효 및 무효 전력이다.here,

Is the total active and reactive power of the load,

Is the active and reactive power of the i bus,

Is the active and reactive power of the kth bus, whose value is known,

Is the active and reactive power of the u-th busbar whose value is unknown.

For each unknown load bus, each coefficient may be equally proportional to the rated capacity of the load and may be represented by Equations 13 and 14 below.

는 부하 전력을 모르는 u번째 모선의 정력 유효전력과 정격 무효 전력이고,

는 부하 전력을 모르는 u번째 모선의 유효 전력 배정계수와 무효 전력 배정계수이다.here,

Is the nominal active power and the rated reactive power of the u-th bus bus, whose load power is unknown.

Is the active power allocation factor and the reactive power allocation factor of the uth bus line whose load power is unknown.

Assuming that the unknown load buses have the same allocation coefficients, Equations 13 and 14 may be represented by one active power allocation coefficient and one reactive power allocation coefficient, as shown in Equations 15 and 16.

는 부하 배분을 위한 유효전력 배정계수 및 무효전력 배정계수이다. only,

Are active power allocation factors and reactive power allocation factors for load distribution.

Using Equation 1, active power allocation coefficients and reactive power allocation coefficients can be obtained as shown in Equations 17 and 18 below.

는 계통의 총 손실의 유효전력과 무효전력이고,

는 변전소의 총 공급전력의 유효전력과 무효전력이며,

는 조상기의 총 전력의 유효전력과 무효전력이고,

는 선로의 병렬 커패시턴스에 의한 총 충전 전력의 유효전력과 무효전력이며,

는 기지(旣知) 부하의 총 전력의 유효전력과 무효전력이고,

는 미지(未知) 부하의 총 전력의 유효전력과 무효전력이다. here,

Is the active and reactive power of the total losses of the system,

Is the active power and reactive power of the total supply power of the substation,

Is the active power and reactive power of the total power of the ancestors,

Is the active and reactive power of the total charging power due to the parallel capacitance of the line,

Is the active and reactive power of the total power of the known load,

Is the active and reactive power of the total power of the unknown load.

Applying the calculated allocation factor to all unknown load buses, the active power and the reactive power can be distributed as shown in Equation 19.

)를 수학식 20을 이용해 계산하고, 전류원에 의한 전압을 상기 수학식 8을 이용해 계산한 후에 각 계산된 전압 결과를 수학식 21과 같이 합성한다. In the above state estimation algorithm performs a tidal current calculation after the step of calculating the load distribution is completed.

) Is calculated using Equation 20, and the voltage by the current source is calculated using Equation 8, and then each calculated voltage result is synthesized as in Equation 21.

Thereafter, if the deviation between the previous value and the updated value of the voltage value is within the allowable deviation, the step of performing the tide calculation is stopped, otherwise the step of performing the tide calculation is repeated.

On the other hand, the step of determining whether the convergence by checking the deviation is calculated by using the equation (10) while satisfying the power supply conditions of the equation 1 with the updated voltage, the loss function before the value and the updated value If the deviation is within the allowable deviation, stop the entire iteration step.

However, if a deviation between the previous value and the updated value of the loss function occurs, it means that it does not meet the power supply and demand conditions, and the iterative calculation step is repeatedly calculated. The initial loss function starts with the value calculated by the voltage given as an initial value as shown in Equation 7, but if it is repeatedly calculated until the deviation of the loss function reaches the allowable level, the power supply and demand conditions can be satisfied. This is because the load power allocation coefficient can be obtained and the unknown load can be allocated.

Hereinafter, the results of the simulation of the state estimation method of the power distribution system according to an embodiment of the present invention in a plurality of embodiments will be described in detail.

In the embodiment of the present invention, the validity and reasonableness of the state estimation algorithm algorithm are examined by performing a simulation by setting some loads as unknown loads.

Table 1 below is divided into data to be used in various embodiments, the verification data of the tide calculation algorithm is to consider the number of loads from the original data of the IEEE 123 node test distribution line as an unknown load and the equivalent load capacity and power factor The input data for the load distribution review is input with arbitrary load capacity and power factor considering some loads as unknown loads from the raw data of the IEEE 123 node test distribution line.

이다. In the case of using the tidal calculation algorithm verification data, since the actual load value is converted into the load capacity and the power factor for the data classified as the unloaded load, the expected load power allocation coefficient is

to be.

First, the validity of the algal calculation algorithm is examined.

(1) Example 1 Data for Algorithm Verification / 1 Unknown Load Application

Table 2 shows the data applied in Example 1, where kW and kVAr shown on the left are actual loads, and kVA and pf shown on the right are equivalent capacity and power factor when set to unknown values.

The load power allocation coefficient when the simulation is performed by applying the data to the right of Table 2 is shown in Table 3, and it can be seen that the load power allocation coefficients are all 1, which results in a reasonable result as expected. In addition, Table 4 summarizes the calculation results, and it can be confirmed that the deviation is very accurate even in the individual voltage deviation.

(2) Example 2 data for algorithm verification / 6 unknown loads applied

Table 5 shows the data applied in Example 2, where kW and kVAr shown on the left are actual loads, and kVA and pf shown on the right are equivalent capacity and power factor when set to unknown values.

In Table 5, the load power allocation coefficient when the simulation is performed by applying the right data is shown in Table 6, and the load power allocation coefficient is 1 for a plurality of unknown loads. If you look at the summary of the calculation results in Table 7, you can see that it is very accurate even in the deviation of individual voltages.

Next, since the capacity and power factor are set to a value different from the actual load value for the unloaded load, the value of the load allocation coefficient at this time is expected to have a different value from the data for verifying the tidal current algorithm. Can be.

(3) Example 3 load distribution review data / 1 unknown load application

Table 8 shows the data applied in Example 3, where kW and kVAr shown on the left are actual loads, and capacity and power factor when kVA and pf shown on the right are set to unknown values.

The load power allocation coefficient when the simulation is performed by applying the right data of Table 8 above is shown in Table 9, and when the load power allocation coefficient of Table 9 is applied to an unknown load, it can be shown as shown in FIG.

As shown in Table 10, when the calculated allocation factor is applied to the unloaded load, the deviation from the actual value is 0.19 kW and 0.18 kVAr, which is very small, resulting in a rational load distribution.

Table 11 summarizes the results of the calculations and shows that they are very accurate even in the deviation of individual voltages.

(4) Example 4 load distribution review data / 6 unknown load application

Table 12 shows data applied to Example 4, where kW and kVAr shown on the left are actual loads, and capacity and power factor when kVA and pf shown on the right are set to unknown values. The fourth embodiment differs from the third embodiment by expanding the unknown load to six.

Table 13 shows the overall ratio of unknown loads.

The load power allocation coefficient when the simulation is performed by applying the data to the right of Table 12 is shown in Table 14.

When the load power allocation coefficient of Table 14 is applied to an unknown load, it is shown in Table 15. When the calculated allocation coefficient is applied to an unknown load, the maximum deviation from the actual value is 3.2 kW and 4.8 VAr. It can be seen that the unknown load of 3 increases slightly compared to the case of one.

However, in the total amount of loads, the deviations are 0.2 kW (300.2-300) and 0.2 kVAr (175.2-175), and it can be seen that the total amount of unknown load distribution is close to the actual total amount with a very small deviation. This is a result of assuming that the allocation coefficients of all the unknown loads are the same. It can be seen that the distribution of the total unknown loads is very reasonably performed although there is a slight variation in the distribution of individual unknown loads.

Table 16 summarizes the calculation results, and it can be seen that it is very accurate even in the deviation of individual voltages.

(5) Example 5 load distribution review data / 15 unknown loads applied

Table 17 shows data applied to Example 5, where kW and kVAr shown on the left are actual loads, and capacity and power factor when kVA and pf shown on the right are set to unknown values. Indicates a ratio.

The load power allocation coefficient when the simulation is performed by applying the right data of Table 17 is shown in Table 19.

When applying the load power allocation coefficient of Table 19 to the unknown load, it is shown in Table 20. When the calculated allocation coefficient is applied to the unknown load, the maximum deviation from the actual value is 7.6kW and 8.2kVAr, and the unknown load is You can see that more increases.

However, in terms of the total load, the deviations were 0.8 kW (728.1-727.3) and 0.9 kVAr (419.8-418.7), indicating that the total amount of unknown load allocation was very close to the actual total amount.

As in Example 4 above, this is a result of assuming that the allocation coefficients of all the unknown loads are the same, and there is a slight variation in the distribution of the individual unknown loads, but the distribution of the total unknown loads is performed reasonably. Can be.

Table 21 summarizes the results of the calculations, and it can be seen that they are very accurate even in the deviation of individual voltages.

It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined in the appended claims. It will be possible.

The present invention relates to a method by estimating the state of a power distribution system, and more particularly, to calculate the load of an unknown bus by reflecting the drawing value of the substation and the known values by measuring some buses, and calculating the current of the power distribution system by using the same. By repeating the process, the load distribution can not only be reasonably calculated and corrected, but also the accurate current calculation algorithm can be used to accurately identify the current state of the distribution system based on the mathematical analysis results to identify current problems and violations. It relates to the estimation of the status of a power distribution system that can be identified and resolved to ensure that it operates reasonably.

Claims (5)

After configuring the Ybus matrix of the power distribution system and setting the system initial value,

An initializing step of calculating; And
An iterative calculation step of performing a flow calculation after performing load distribution based on state estimation, and determining whether or not convergence by checking a deviation;
Method by the state estimation of the power distribution system, characterized in that consisting of. The method of claim 1,
Setting the system initial value,
a) setting the injection power of the substation as the total supply power,
b) setting the voltage of the load busbar to 1.0pu as the voltage of the substation is known;
c) setting the unknown load power to the load capacity;
d) A method of estimating a state of a power distribution system, comprising: setting an active power allocation coefficient and a reactive power allocation coefficient to 1 for load distribution. The method of claim 1,
Performing the load distribution,
a) power of the unknown load using the equation

),

only,

Is active power allocation factor and reactive power allocation factor for load distribution,

Is the nominal active power and nominal reactive power of the uth bus without knowing the load power.
b) calculating bus power using the following equation;

c) calculating a loss function using the following equation;

only,

Is the active and reactive power of the total losses of the system.
d) calculating the active power allocation coefficient and the reactive power allocation coefficient using the following equation;

Only,

Is the active power and reactive power of the total supply power of the substation,

Is the active power and reactive power of the total power of the ancestors,

Is the active and reactive power of the total charging power due to the parallel capacitance of the line,

Is the active and reactive power of the total power of the known load,

Is the active and reactive power of the total power of the unknown load.
e) the power of the unknown load (

Method of estimating the state of the power distribution system, characterized in that consisting of a) step. The method of claim 1,
Performing the algal calculation,
a) injection current of the load bus

),
b) voltage by current source (

),
c) synthesize the voltage results (

),
d) estimating the tidal current calculation step if the deviation between the previous value and the updated value of the voltage value is within the allowable deviation, and repeatedly calculating the tidal current calculation step if it is not within the allowable deviation; Way. 5. The method according to any one of claims 1 to 4,
Determining whether or not convergence by checking the deviation, after calculating the loss function with the updated voltage (V), if the deviation between the previous value of the loss function and the update value of the loss function is within the allowable deviation and stops the entire step; If it is not within the allowable deviation, it is repeatedly calculated until the deviation of the loss function reaches an acceptable level.

Images