JPH02123944A - Estimation of winding temperature - Google Patents

Abstract

PURPOSE:To perform the estimation of temperature by obtaining a first order lag temperature characteristic by a winding current value and then by further obtaining the first order lag characteristic by the variations of the winding current value in addition to the said temperature characteristic. CONSTITUTION:Thermal input and temperature output response will be in the first order lag system. Firstly, in the range of time O<=t<a if the motor current I1 is caused to flow, the temperature rise characteristic of a motor winding temperature thetau1 will be obtained. If the motor current is changed to I2 in the range of time a<t<=b, the motor winding temperature thetaD1 will show the falling characteristic of the first order lag in the current difference (I1-I2). Furthermore, when the rising characteristic by the current I2 after the current change is added and overlapped, the temperature falling characteristic is obtained. Moreover, if the motor current I3 is caused to flow in the range of time b<t, the motor winding temperature thetau2 will show the rising characteristic of the first order lag in the current difference I2-I3, which is added and overlapped by the rising characteristic by the current I2 and becomes the temperature rising characteristic.

Description

DETAILED DESCRIPTION OF THE INVENTION ^ Industrial Application Field The present invention relates to a winding temperature prediction method for maximizing the thermal use of equipment such as motors.

B. Summary of the Invention ゛The present invention uses a thermal energy balance formula, that is, the amount of heat released minus the amount of heat released, in order to allow equipment that adjusts the winding current, such as motors, to operate without failure even under severe usage conditions exceeding the rating. relates to the temperature change of the device, and by discovering that the temperature change in this equation has a first-order delayed response, we have made use of this to prevent burnout or failure even if the device is used to the point of overheating. be.

C1 Conventional technology and its problems Conventionally, for devices with @ wires, such as motors, there is winding protection based on the winding temperature, and the simplest measure is to cut off the power when the upper limit temperature is exceeded. be. However, in actual operation, there may be cases where it is necessary to avoid stopping the motor at all costs; for example, when operating the manipulator, there may be cases where it is desired to reduce the load on the motor and continue the operation without stopping.

For this purpose, it is necessary to detect the winding temperature and predict temperature changes, but even if temperature detection based on current values is possible, temperature prediction is difficult due to a time delay.

In particular, when the motor is used under conditions beyond its rated rating, it is extremely difficult to predict the temperature since it is not used normally.

In view of the above-mentioned problems, the present invention provides a winding temperature prediction method that predicts t2Hr degree during continuous operation of a device such as a motor, or when the device is used at or above the rated value.

Means for Achieving the Problem The present invention achieves the above-mentioned object by discovering that a one-grade change results in a first-order delayed response, and by applying the fact that this delayed response can be considered based on the principle of superposition. Based on Party B.

A first-order lag temperature characteristic is obtained by the action winding current value, and then a first-order lag temperature characteristic is obtained by changing the winding current value in addition to the above-mentioned temperature characteristics, and further by a change in the winding current value. By adding the previous temperature characteristics, it became possible to predict the temperature more accurately than the load current history.

Embodiments Here, embodiments of the present invention will be described with reference to the drawings. For example, first consider the basic heat balance equation for a motor. Now Q is the amount of heat generated (proportional to the square of the armature current), H is the heat radiation coefficient (proportional to area, etc.), M is heat capacity (related to shape),
If θ is the temperature, the following equation is obtained.

In other words, this equation shows that the temperature rise of a device having a heat capacity M is obtained by subtracting the amount of heat released from the heat generated (1).

Although we would like to obtain the temperature θ based on the above equation, the heat input (
Assuming that the Joule loss proportional to the square of the current and the temperature output response are a first-order lag system, for example, Fig. 1 (al fb
Calculations can be made for a certain model as shown in Figure 1. That is, first, when the motor current 11 is passed in the range of time O≦b≦a, the motor winding temperature θul becomes r', (1-
e-〒)・Improves the temperature rise characteristic expressed by equation (1).

Note that here, y,, r'', , is the temperature rise resistance when I flows for an infinite time, and T1 is the specific number at which the temperature rises.

In addition, when the motor current is changed to - within the range of time a (t≦b), the motor winding temperature θ.1 (2 equations In this equation, the point θ. The current I2 after the current change shows the falling characteristic of the first-order lag at
By adding and superimposing the rising property due to humidity, the humidity falling property is obtained. Note that here the two bows are I and ga I. T2 is the time constant when the temperature decreases from 0 to 0.

Furthermore, when the motor current I flows in the range of time bat, the motor winding temperature 0.2 is calculated by the following formula... This is a temperature increase characteristic that is superimposed by taking into account the increase characteristic due to the current I.

In this way, the temperature at the time of the current change, the rising characteristics due to the current before and after the change, and the rising or falling characteristics of the difference in the current before and after the change, are superimposed, and 1
This means that the temperature can be predicted by combining the second-order lag characteristics.

Next, consider the period when calculating equations (1), (2), and (3) above.In other words, when considering the gain in a first-order lag system with temperature characteristics, a decrease in gain means that the frequency increases. Since it becomes large, this means that it is sufficiently attenuated.

here, The temperature characteristics are expressed by the following equation with a first-order lag.

That is, in Fig. 2, the input heat flow and the output temperature are first-order lags, and in this case T differs depending on the motor, but in the example where T = 10 minutes, lGθ. .

Let's assume that it is ω of □.

This formula has 10,000 degrees ω (2π times 10·60) 2, that is, the calculation cycle is only about 30 seconds and the characteristics are sufficiently reflected.

Considering that the average temperature within this calculation cycle is, for example, full speed in 1 second, it is as shown in Figure 3 (a) (500 m
The current is measured every s, and the average value is calculated for 30 seconds.

That is, instead of the instantaneous current 10., the value ■ is obtained by averaging these values within the calculation period as shown in the following equation. ie seek.

■==-Miyu 迭'' 42L B2 供侃6) Equation Here, 1 is the current at the time of sampling, I is the average current, and n is the sample cycle.

FIG. 3 (bl is the motor winding temperature based on the current shown in FIG. 3 (al).

The temperature θ can be obtained by obtaining each characteristic of temperature rise or 1 degree fall using the calculation cycle as described above. can be found. FIG. 4 shows the processing flow. That is, first of all, the above (
6) Find ■ using the formula. And of the previous calculation cycle!
. Compare -1 and ■.

If T, ≧Il'l-1, move to the rising characteristic step, and again ■. If <■,, -□, move to the descending characteristic step and θ at any point. Find out. That is, 1, ≧1. ．．
−． Then, the temperature θ is at tn-1. -1 condition, I',
−.

(r,,-1,)2 energy is input in a step manner and superimposed on the energy of I,,<1. , then the energy of 1■ knee 1 and -1 (1,, -, -[,)
The energies of 2 are input in steps and superimposed.

Next, the external temperature T is added to the obtained θ. Find the temperature θ' by adding .

Simulation> Here, a simulation was performed using a model motor based on temperature rise experimental data and compared with actual measured values.

The model shown in FIG. 5 shows the temperature rise in the part of winding i when the rated current is passed in a restrained state. Here, 2.25
During continuous operation at A, the temperature rose until it reached 168 degrees (the temperature increase relative to the rated current was 148 degrees), 63
%, the thermal time constant was 25.2 minutes.

The model was approximated as a first-order lag system represented by these values.

Figure 6 shows the case where a current equivalent to 148 deg when applied continuously is intermittently input every 10 minutes, and the motor rated current is 2.
This is a comparison between the experimental temperature rise value corresponding to 25A and the ξ shiureshi date, and the movements are very similar at approximately constant intervals.

Figure 7 shows the current of 1.5 @ 3.37 with a rated current of 2.25 A.
This is a comparison between actual measurements and simulation results obtained when a similar experiment was conducted with 5 A flowing, and where a current equivalent to 333 degrees was intermittently input every 10 minutes. Here, it is obtained from 333 deg=148 x (1°5)2. In the case of FIG. 7 as well, the movements are similar at approximately constant intervals vertically.

8 and 9 are obtained by performing the same stain correction scenes in consideration of correction coefficients so that the peak values in FIGS. 6 and 7 match the simulated values and the actual values. As a result, it can be seen that even the rated and 1.5 times the rated loads use the same correction coefficient and are in good agreement. Furthermore, the reason that there is an error in the rising portion in both FIGS. 8 and 9 is thought to be due to the rising portion of the actual measured values being somewhat different from the simulation conditions, and it is thought that the accuracy will improve if the same experiment is performed. In addition, the correction coefficient is KX (2, 2, 5) 2=
148de (then = 29.23, and this K was set as 24.0. In other words, 24X (2, 25)"
Correct the current to be equivalent to =121.5deg.

As described in detail, according to the present invention, it is possible to predict the temperature rise of a device having windings, and when it comes to motors, it is possible to use them without failure up to the very limit of their use. It is also possible to predict the winding temperature of motors that do not have a winding thermometer, which can be used as an algorithm for motor protection.

[Brief explanation of the drawing]

Figure 1 is an explanatory diagram of motor winding temperature according to the first-order lag related to motor current, Figure 2 is an explanatory diagram of input/output of motor temperature (heat), and Figure 3 is a diagram of specific motor current and motor winding. Figure 4 is a graph showing the actual measurement of the motor winding temperature of the model. Figure 6 is a graph showing the actual measurement and simulation results at the rated motor current. , Figure 7 is a graph showing actual measured simulation results at a current 1.5 times the rated value, Figures 8 and 9.
The figures are graphs showing actual measured values and simulation results in consideration of the correction coefficients in FIGS. 6 and 7, respectively. In the figure, θ. 1. θ,,2. θ01'θn-1'θ. is the temperature characteristic, θ′ is the temperature characteristic considering the external temperature, I,, +2 , I,, I,,,
I,,f,t=flow, T is the time constant when the temperature rises, T is the time constant when the temperature falls.

Claims (1)

[Claims] Measure the winding current value multiple times within the calculation cycle, calculate the average value, compare this average value with the average value in the previous calculation cycle, and determine whether the current average value is large or small. KI to
A volume characterized in that a temperature change characteristic is calculated based on a Joule loss of ^2 and a temperature response with a first-order lag of (1-e^-^t^/^T), and the external temperature is further taken into account in this characteristic. Line temperature prediction method. Note that K is a coefficient, I is a current value, T is a time constant, and t is a time.