JP2016011869A - Metering device and metering method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To achieve a highly accurate measurement and a shorter measurement time.SOLUTION: A metering device includes: a metering part 2 which outputs electric information corresponding to weight of a measuring object 3; a data holding part 10 which acquires the electric information as measurement data; an adaptation processing part 11 which smooths the measurement data by a moving average, calculates adaptation moving average number which is the moving average number minimizing variation of the moving average, and uses it for the smoothing; and an indicator 9 which displays the value smoothed by the moving average as measured value.

Description

  The present invention relates to a weighing device and a weighing method.

  In the field of manufacturing industry and distribution, measuring devices using load cells have been widely used. The load cell has a strain gauge bonded to an elastic body, and is configured to obtain a voltage output corresponding to the amount of strain of the elastic body generated when a load is applied. This output value is amplified and then digitally converted, and a measured value is calculated based on the digitally converted output value. Such a weighing device is configured such that when the object to be weighed is placed on the loading platform of the weighing unit, a load is applied to the load cell, whereby the weighing value of the object to be weighed is displayed. Immediately after the object to be weighed is placed on the loading platform, a free damping vibration occurs and the displayed value is gradually stabilized. Moreover, the vibration at this time vibrates with a period close to the natural vibration of the spring-mass system, and the period varies depending on the weight of the object to be weighed on the loading platform. On the other hand, in order to quickly stabilize the oscillating measurement value, it is known that it is effective to average over one time of the vibration period or a time width that is an integral multiple of the vibration period. Several proposals have been made.

  Patent Document 1 discloses a measuring device including an integrator that sets a time width (τ) obtained by calculation and a time width that is an integral multiple of the time width and uses this time width as an integral time width. However, when the weighing device is a relatively small device such as a commercial desktop digital weighing device, the spring constant of the load cell and the mechanical elements in the weighing device connected to it can be grasped at the design and development stage of the manufacturer. Although it is possible to obtain the period by calculation, in the case of a relatively large industrial weighing device, the vibration period is greatly influenced by the support part at the installation location, and the spring constant of the support part is unknown, and the vibration is calculated. It is often difficult to determine the period.

  Patent Document 2 discloses a method in which a vibration period is directly obtained by a Fourier transformer using a load cell balance having a support structure with a suspended structure, and the obtained period and the averaging time by moving average are matched. However, the Fourier transform can divide the periodic function into vibration functions and obtain the spectral components for each frequency. The output result of (Fast Fourier Transform) includes a low frequency spectrum component in addition to the natural vibration period to be obtained, and it is difficult to extract an effective period. In addition, as a component device, a DSP (Digital Signal Processing) dedicated to FFT and a high-performance CPU (Central Processing Unit) are required, which is expensive.

  In Patent Document 3, as another method for obtaining the vibration period, the peak value of the vibration waveform and the time interval (mountain-mountain or valley-valley) of the peak are detected, and an average value between them is calculated and output as a measured value. A display method has been proposed. However, if there is a high-frequency or low-frequency noise component, it is difficult to detect a peak, and a stable time interval and an average value between them cannot be obtained, and a stable measurement value may not be obtained.

Japanese Patent Laid-Open No. 62-261021 JP 11-311566 A JP 59-54931 A

  An object of the present invention is to provide highly accurate weighing and shortening of the weighing time.

  According to a first aspect of the present invention, there is provided a weighing unit that outputs an electrical signal corresponding to the weight of an object to be weighed, a data holding unit that takes in the electrical signal as weighing data, and smoothing the weighing data by moving average, An adaptive moving average number that is a moving average number that minimizes fluctuations in the moving average is calculated, and an adaptation processing unit that is used for the smoothing and a display that outputs the value smoothed by the moving average as a measured value A weighing device is provided.

  According to this weighing device, since smoothing is performed by moving average using the adapted moving average number, it is possible to measure with high accuracy and shorten the weighing time. Here, the moving average number represents the number of data used when calculating the moving average value at each time. In general, when the vibration of the measurement data takes a moving average with a natural vibration period or an average time width that is an integral multiple of the natural vibration period, the amplitude is almost zero. In other words, the moving average value in this case becomes the measured value after the vibration is converged. Therefore, in this case, the weight value after convergence can be obtained accurately without waiting for the convergence of the vibration of the weight data. In this weighing device, the moving average number is changed, and the moving average number corresponding to the natural vibration period or an average time width that is an integral multiple thereof is calculated as the adapted moving average number, thereby minimizing the fluctuation of the moving average. . Therefore, highly accurate weighing and shortening of the weighing time are possible.

  It further includes an operation setting unit capable of inputting and setting information necessary for calculation in the adaptation processing unit from the outside, and the adaptation processing unit sets the moving average number set in advance or input by the operation setting unit Is changed to a continuous integer of 3 or more including the moving average number in between, and the moving average of the weighing data is calculated by trial, and a variation evaluation value representing a variation of the moving average in each moving average number is respectively calculated. It is preferable to calculate and determine the adapted moving average number based on each of the fluctuation evaluation values.

  According to this weighing device, the moving average is calculated with a plurality of continuous moving average numbers, the fluctuation evaluation value for each moving average is obtained, and the magnitude of these is compared to reliably minimize the fluctuation of the moving average. The moving average number (adapted moving average number) can be obtained.

  The moving average is a single moving average or a double moving average, and the fluctuation evaluation value is obtained by taking a difference between data adjacent in the time lapse direction with respect to the time series data of the moving average, and calculating the square of the difference. The sum is preferably calculated.

  According to this weighing apparatus, by setting the fluctuation evaluation value as described above, the fluctuation (amplitude) of the moving average can be numerically evaluated with ease. For the fluctuation evaluation value, a deviation sum of squares used in statistical processing may be used. However, when the time series data includes monotonic increase or low frequency, it is affected by these. Since the fluctuation evaluation value does not use the concept of the difference from the average value unlike the deviation sum of squares, it is more advantageous than the deviation sum of squares in that it is not easily affected by monotonic increase or low frequency.

  The moving average is a double moving average, and the fluctuation evaluation value is obtained by taking a difference of data at the same sampling time with respect to time series data of the double moving average of adjacent moving average numbers. The sum of squares may be calculated.

  According to this measuring device, by setting the fluctuation evaluation value as described above, the fluctuation (amplitude) of the moving average can be converted into a numerical value and evaluated in the same manner. The double moving average is characterized in that the amplitude is smaller in the vicinity of the moving average number where the amplitude is zero compared to the single moving average. Therefore, when the double moving average is used, since the amount of fluctuation is small compared to the single moving average, stable weighing is possible. In the case of the double moving average, when the natural vibration period is τ, the number of moving averages coincides with the average time width of 2τ, 4τ,. Further, in the vicinity of 2τ, 4τ,..., A downwardly convex curve is formed. Therefore, the difference in amplitude corresponding to the adjacent moving average number increases as the distance from 2τ, 4τ,. By utilizing this characteristic and evaluating the difference in amplitude corresponding to the number of adjacent moving averages, the fluctuation of the moving average can be evaluated.

  After calculating the adaptation moving average number, the storage unit stores the relation data of the adaptation moving average number and the corresponding metric value, and stores the relation data every time the measurement is repeated, and further includes the adaptation The processing unit derives a relational expression between the adaptive moving average number and the measurement value based on the relational data, and uses the relational expression according to the mode set in the operation setting unit when newly measuring. It is preferable to determine the adapted moving average number and update the moving average number.

  According to this weighing device, since the adapted moving average number corresponding to the measured value can be obtained immediately, it is possible to prevent a response delay due to the adaptation processing time as compared with the case of performing the normal moving average adaptation processing.

  According to a second aspect of the present invention, when the amplitude of the weighing data is equal to or greater than a predetermined value and the change rate of the weighing data is equal to or less than another predetermined value, the weighing data is smoothed by a moving average, The moving average number of the moving average is changed to at least 3 or more continuous integers including the moving average number, and the moving average of each moving average number is calculated by trial, and the fluctuation of each moving average is calculated. Each representing a fluctuation evaluation value to be represented, determining an adapted moving average number that is the moving average number that minimizes the fluctuation of the moving average based on each of the fluctuation evaluation values, and converting the metric data into the adapted moving average A weighing method is provided in which a value smoothed by the moving average of the number is used as a weighing value.

  According to this weighing method, since smoothing is performed by moving average using the adapted moving average number, it is possible to measure with high accuracy and shorten the weighing time. Calculate the moving average with multiple consecutive moving average numbers, and compare the magnitudes of the fluctuation evaluation values for each moving average to obtain the moving average number (adapted moving average number) that minimizes the fluctuation of the moving average. Can do. By smoothing the weighing data by the moving average of the adapted moving average number, the weighted value after convergence can be accurately obtained without waiting for the convergence of the vibration of the weighing data.

  According to the present invention, since smoothing is performed by moving average using the adapted moving average number, it is possible to perform highly accurate weighing and shorten the weighing time.

The conceptual diagram which shows the measuring device of this invention. The figure which shows the time change of change rate per vibration 1 period of measurement data, smoothing output, and measurement data. The conceptual diagram which shows a single moving average and a double moving average. Explanatory drawing which shows the time change of single moving average process data. Explanatory drawing of the smoothing characteristic by a single moving average and a double moving average. Explanatory drawing which shows the time change of the difference of single moving average process data. The figure which shows the fluctuation | variation evaluation value in each moving average number of FIG. The flowchart which shows a measurement process program. 8B are two sub-flowcharts of the moving average number adaptation process that is a part of FIG. 8A. The conceptual diagram which shows the measuring device of 2nd Embodiment of this invention. The flowchart which shows a part of calculation processing program of 2nd Embodiment of this invention. Explanatory drawing which shows the relational expression of the adaptation moving average number and measurement value. The figure which shows the track scale which is an example of application of this invention. The figure which shows the crane scale which is an example of application of this invention. The figure which shows the vehicle-mounted measurement which is an example of application of this invention.

  Next, embodiments of the present invention will be described with reference to the accompanying drawings.

(First embodiment)
FIG. 1 shows a weighing device 1 according to a first embodiment of the present invention. The weighing device 1 measures the weight of the object 3 placed on the weighing unit 2.

  The weighing device 1 includes a weighing unit 2, a support unit 4, an amplifier 5, an A / D converter 6, a control calculation unit 7, an operation setting unit 8, and a display 9.

  The weighing unit 2 is a part on which the object to be weighed 3 is placed. The weighing unit 2 has a load cell 2a. The load cell 2a outputs a voltage in accordance with the amount of strain generated when a load is applied to the measuring unit 2. The support part 4 supports the measuring part 2. In the present embodiment, a method of placing the object 3 to be measured on the weighing unit 2 is adopted, but other methods such as a hanging method may be used, and the mode of the weighing unit 2 is not particularly limited.

  The output voltage of the load cell 2a is transmitted to the amplifier 5 as an electric signal. The amplifier 5 amplifies this electric signal. The amplified electrical signal is transmitted to the A / D converter 6. The A / D converter 6 converts this electrical signal from an analog signal to a digital signal. The digital signal converted by the A / D converter 6 is transmitted to the control calculation unit 7.

  The control arithmetic unit 7 is constructed by hardware including a storage device such as a CPU (Central Processing Unit), a RAM (Random Access Memory), and a ROM (Read Only Memory), and software installed therein. The control calculation unit 7 includes a data holding unit 10 and an adaptation processing unit 11. The adaptation processing unit 11 includes a smoothing processing unit 12.

  The data holding unit 10 takes in the signal transmitted from the A / D converter 6 as measurement data at a predetermined sampling interval.

  The operation setting unit 8 inputs and sets the initial value of the moving average number N and the operation mode. For example, the operation mode includes four modes of operation modes 0 to 3 and can be selected by the operation setting unit 8. In the operation mode 0, the adaptation process is not executed, and the input value of the operation setting unit 8 is used as the moving average number. In the operation mode 1, the adaptation process is executed, and the moving average number update is performed after the adaptation process is completed, and the moving average number set in advance by the adaptation processing unit 11 is set when the next measurement is started. In the operation mode 2, the adaptation processing is executed, and the moving average number update is performed after the next measurement is started, and becomes the moving average number set in advance by the adaptation processing unit 11. In the operation mode 3, the adaptation process is executed, and the moving average number update is immediately applied with the adapted moving average number corresponding to the measured value from the accumulated data after the start of the measurement, and the adaptation processing unit 11 performs the preset movement. Average number. When executing the adaptation process, the adaptation process is performed only once for each measurement.

  The adaptation processing unit 11 acquires the measurement data from the data holding unit 10. The measurement data is smoothed by the moving average in the smoothing processing unit 12. Further, the adaptation processing unit 11 executes an adaptation process for the moving average number N, and determines the adapted moving average number Na. The adaptation processing unit 11 may not execute the adaptation process of the moving average number N depending on the operation mode and the operation situation.

  The display unit 9 displays the measurement data smoothed by the moving average after the adaptation process or the measurement data that has not been subjected to the adaptation process as a measurement value. The user recognizes the weight of the object to be weighed 3 placed on the weighing unit 2 by confirming the display on the display unit 9.

  Next, the principle of the weighing method executed by the weighing device 1 will be described.

  When the object to be weighed 3 is placed on the weighing unit 2, the weighing value is not stabilized due to the bending of the weighing unit 2 or the support unit 4, and damped free vibration occurs. FIG. 2 shows metric data that oscillates damped, a smoothed output obtained by smoothing the data, and a rate of change per cycle. The rate of change here is the rate of change per cycle of vibration. Therefore, when the vibration is in a steady state, this rate of change is almost zero. The weighing device 1 can acquire an accurate measured value without waiting for the convergence of the vibration.

  The vibration period of the measured value varies depending on the weight of the object 3 to be weighed. For the sake of explanation, let τ be the vibration period of the measured value. The sampling period of the weighing data after being digitized by the A / D converter 6 is assumed to be Δt. It is known that the influence of vibration is most reduced in principle when the weighing data is smoothed by moving average with N (moving average number) × Δt = τ. The present invention uses the smoothing characteristic of this moving average. Trial calculation is performed with a plurality of continuous moving average numbers N, and a moving average number with a small vibration influence is obtained by comparing fluctuation evaluation values corresponding to deviation sum of squares. In this way, the moving average number is adapted following the vibration cycle that changes depending on the weight of the object 3 to be measured, and the measurement time can be shortened and the measurement can be performed with high accuracy.

  The moving average used in the present embodiment is a double moving average obtained by applying a simple moving average (single moving average) and a simple moving average twice. Assuming that the time series weighing data is {Xi} and the moving average number is N, the single moving average S [N, i] and the double moving average D [N, i] at time i are expressed by the equations (1) and (1), respectively. It is expressed as (2).

Ｎ：移動平均個数
Ｘ：計量データ

N: Moving average number
X: Weighing data

Ｎ：移動平均個数
Ｓ：１重移動平均

N: Moving average number
S: Double moving average

  FIG. 3 is a conceptual diagram when the moving average number N = 5 for the single moving average and the double moving average. The single moving average repeats averaging of the past five measurement data at each time. The double moving average repeats averaging of the past five pieces of measurement data at each time in the same manner with respect to the single moving average result.

  In order to show the smoothing characteristics by the vibration period and moving average, an example of the response calculation result when the moving average number N is changed with a sine wave (amplitude A0 = 10 kg, period τ = 400 ms) and a sampling period Δt = 40 ms is used. To explain.

  FIG. 4 shows a time change of the single moving average S [N, i]. A transient response is shown for approximately τ (sec) immediately after the start of the moving average, but thereafter a steady state is obtained. In particular, at N = 10 where N × Δt = τ, the fluctuation amplitude becomes zero.

  FIG. 5 is a diagram showing the smoothing characteristic by the moving average. FIG. 5 shows the relationship between the amplitude after moving to the steady state and the moving average number, with the single moving average and the double moving average arranged side by side. In the single moving average, the amplitude becomes zero at τ, 2τ, 3τ, 4τ. In the double moving average, the amplitude is zero with the moving average number matching the average time width of 2τ, 4τ. In addition, the amplitude decreases with the moving average number in the vicinity, and increases with distance.

  In comparison between the single moving average and the double moving average, for example, the amplitude of the double moving average is smaller in the moving average number in the vicinity of 2τ where the amplitude is zero. Therefore, a more stable measurement value can be expected. Calculate the moving average with a number of continuous moving averages, and if the fluctuation is considered to be a variation, calculate the sum of squared deviations used in statistical processing, and compare the magnitudes of these values to find the moving average with the smallest fluctuation I can do it. An equivalent method will be described below.

  Usually, the deviation sum of squares calculates the deviation for each data from the average value (or predicted value) of the sample data, and takes the sum of the squares. Here, the average value is not used, and the sum of squares of the differences between the data adjacent in the time lapse direction is calculated, and this is set as the fluctuation evaluation value P [N]. That is, in the case of a single moving average, it is expressed as the following formula (3).

Ｎ：移動平均個数
Ｓ：１重移動平均

N: Moving average number
S: Double moving average

  Similarly, in the case of the double moving average, it is expressed as the following formula (4).

Ｎ：移動平均個数
Ｄ：２重移動平均

N: Moving average number
D: Double moving average

  A normal deviation sum of squares may be used as the fluctuation evaluation value, but this is calculated by calculating the deviation from the average value. Therefore, when the time series metric data includes monotonic increase or low frequency, it is affected by this. Compared with this, in the case of Formula (3) and Formula (4), there exists an advantage which is hard to be influenced.

  By comparing the magnitudes of the respective P [N] values, the moving average number N having a smaller value can be obtained as the adapted moving average number Na. The difference value of the single moving average in the calculation example of FIG. 4 is shown in FIG. When N = 10, the difference of the single moving average is zero, and the measured value does not change. The further away from N = 10, the larger the difference and the metric value fluctuates. Further, FIG. 7 shows a fluctuation evaluation value P [N] obtained by calculating the sum of squares of the differences by the equation (3). The values of P [N] are compared. In this case, N = 10 is minimum, and N = 10 is determined as the adapted moving average number Na.

  The fluctuation evaluation value may be other than the formula (3) and the formula (4). Referring to the smoothing characteristic graph by the moving average of FIG. 5, the curve indicating the smoothing characteristic in the case of the double moving average becomes zero with the moving average number matching the average time width of 2τ, 4τ. Moreover, it becomes a downward convex curve in the vicinity of 2τ, 4τ. Therefore, the difference in amplitude corresponding to the adjacent moving average number increases as the distance from 2τ, 4τ. In consideration of this characteristic, Expression (5) similar to Expression (3) and Expression (4) is defined. The moving average in this case is a double moving average. As the fluctuation evaluation value, the difference between data at the same sampling time is calculated with respect to time series data of double moving averages of adjacent moving average numbers, and the sum of squares is calculated.

Ｎ：移動平均個数
Ｄ：２重移動平均

N: Moving average number
D: Double moving average

  Equation (5) decreases with the moving average number close to the average time width of 2τ, 4τ,. By comparing the magnitudes of the respective P [N] values, the moving average number N having a smaller value can be obtained as the adapted moving average number Na. Further, after each P [N] value is appropriately weighted, the magnitudes may be compared. For example, if a1, a2,... Are weight coefficients, a1 × P [N−2], a2 × P [N−1], a3 × P [N], a4 × P [N + 1], and a5 × P [N + 2]. You can compare the magnitude of values. In this way, if it is difficult to determine the size of each P [N], select the smaller moving average number and place importance on responsiveness, or select the larger moving average number. The weighting can be changed depending on whether the stability of fluctuation is important.

  Next, an operation method of the weighing device 1 will be described.

  FIG. 8A is a flowchart showing the operation of the weighing device 1 according to the first embodiment of the present invention. The data holding unit 10 acquires the measurement data {Xi} from the load cell 2a of the measurement unit 2 via the amplifier 5 and the A / D converter 6 (Step S1). Whether or not the moving average number needs to be updated is determined based on the operation mode and the operation state input and set in advance by the operation setting unit 8 (step S2). For example, if the input value of the operation setting unit 8 is changed in the operation mode 0, the moving average number is immediately updated. In addition to the moving average number input in the operation setting unit 8, a preset moving average number may be used. The previously set moving average number used here includes not only the value set as the initial value of the program but also the adapted moving average number determined in the last adaptation process. If it is determined in step S2 that the moving average number needs to be updated, it is updated in step S3. If it is determined in step S2 that the moving average number update is unnecessary, the process proceeds to step S4. The first time is set to the initial value of the moving average input and set by the operation setting unit 8 (step S3). A single moving average is calculated based on the equation (1) using the set moving average number as the time series weighing data {Xi} (step S4). Thereafter, a double moving average is calculated based on the equation (2) (step S5). Next, the rate of change per predetermined time of the weighing data is calculated (step S6). The predetermined time can be an integral multiple of the sampling time Δt, but is preferably a value close to the vibration period τ, and may be N (moving average number) × Δt. Subsequently, it is determined whether or not the moving average number adaptation is necessary according to the operation mode and operation state of the operation setting unit 8 (step S7). For example, in the operation mode 1, if the adaptation process has not been performed once, the process proceeds to the adaptation process.

If it is determined in step S7 that the moving average number adaptation is unnecessary, whether or not the weighing data smoothed in steps S4 and S5 needs to be further smoothed is determined based on the digital filter constant setting in the operation setting unit 8. (Step S8). For example, 0 to 99% is used as the exponential smoothing filter coefficient. The larger this value, the greater the smoothing effect. When 0% is set, there is no smoothing effect and the branch is N. The filter coefficient is adjusted and set by the operator in view of the stable state of the measured value.
If it is determined in step S8 that smoothing is necessary, the smoothing processing unit 12 performs smoothing with a digital filter such as an exponential smoothing filter (step S9). If it is determined in step S8 that smoothing is unnecessary, the process proceeds to step S10. Then, it is displayed and output as a measured value on the display 9 (step S10).

  If it is determined in step S7 that moving average number adaptation is necessary, collection of measurement data for a certain period is started as preparation for moving average number adaptation processing. FIG. 8B is two sub-flowcharts of FIG. 8A showing the moving average number adaptation process. First, it is determined whether the metric data {Xi} is equal to or greater than a predetermined amplitude threshold value and the change rate of the metric data obtained in step S6 is equal to or less than another predetermined threshold value (steps S11 and S12). As a result, the object 3 is placed on the weighing unit 2, and it is determined whether the measured value is increasing or is the damped vibration after loading (see FIG. 2). When this condition (step S11, step S12) is satisfied, the data holding unit 10 acquires measurement data {Xi} for a certain period (step S14, step S15). When this condition (step S11, step S12) is not satisfied, the process proceeds to step S8. If the acquisition of the measurement data {Xi} for a certain period is not completed even after a predetermined time has elapsed, this process is interrupted and a timer is set so as to proceed to step S8 (step S13). In step S15, when acquisition of the weighing data {Xi} is completed, the adaptation processing unit 11 starts the moving average number adaptation process (the process proceeds to step S21). If acquisition of the weighing data {Xi} is not completed in step S15, the process proceeds to step S11.

  For the weighing data {Xi}, the moving average number N currently in use is changed to a continuous number (for example, N−2, N−1, N, N + 1, N + 2) including the moving average number N in between, In the smoothing processing unit 12, for each moving average number, a single moving average calculation based on the equation (1) and a trial calculation of a double moving average calculation based on the equation (2) are performed (steps S21 to S24). Then, the process proceeds to step S22 until the trial calculation is completed, and these are repeated (step S25). When the trial calculation of the moving average with a plurality of continuous moving average numbers is completed (step S25), the adaptation processing unit 11 uses the obtained smoothed data to evaluate the variation for different moving average numbers N. The value P [N] is calculated using any one of formulas (3) to (5) (step S26). A comparison of the magnitudes of the obtained P [N] is performed (step S27), and the moving average number N of P [N] that is the minimum is obtained. This moving average number N is determined as the adapted moving average number Na (step S28). For example, if P [N-1] is minimum, it is determined that Na = N-1. The timing at which the adapted moving average number Na is updated with respect to the currently used moving average number depends on the operation mode set in the operation setting unit 8 and is determined in step S2. It can be changed immediately after the moving average number adaptation processing is completed, or can be changed from the start of the next measurement. The flowcharts of FIGS. 8A and 8B are explanatory diagrams in the case of using the double moving average. However, the present invention can be performed using only the single moving average. In this case, the double moving average calculation is omitted. it can.

  As described above, since the weighing device 1 performs smoothing by moving average using the adapted moving average number, even in a use environment in which the natural vibration period changes depending on the weight of the object 3 to be measured, The weighing time can be shortened. In addition, it is difficult to be affected by peripheral disturbance vibrations, and stable measurement is possible.

(Second Embodiment)
FIG. 9 shows the weighing device 1 of the second embodiment. The weighing device 1 according to the present embodiment is the same as the first embodiment in FIG. 1 except for the portion related to the storage unit 13. Therefore, the same parts as those shown in FIG.

  In the present embodiment, the control calculation unit 7 includes a storage unit 13. The storage unit 13 stores and stores relational data between the adapted moving average number Na calculated at the time of measurement so far and the measured value W at that time. That is, this relationship data is accumulated every time measurement is repeated.

  In this embodiment, in the final step (step S29) of the sub-flowchart of the moving average number adaptation process in FIG. 10, the metric value W based on the adapted moving average number Na and the smoothed data obtained in steps S21 to S24. Are stored in the storage unit 13. Each time measurement is repeated, this relational data is accumulated, and a relational expression (approximate expression) between the measurement value W and the moving average number N as shown in FIG. 11 is derived from the accumulated data. Using this relational expression (approximate expression), the moving average number for the measured value W is determined. In step S3, a function of changing the moving average number is added. A linear approximation equation or a curve approximation equation is derived from the set of accumulated data (Wi, Ni) using the least square method. In this embodiment, when the weighing object 3 is placed on the loading surface of the weighing unit 2, the adapted moving average number corresponding to the measured value is immediately applied, so that there is a response delay due to the moving average number adaptation processing time. There is no advantage.

  In comparison between the single moving average and the double moving average, the single moving average has a smaller response delay, whereas the double moving average has a characteristic that fluctuations are more stable. Considering these, it can be properly used according to the purpose of measurement. This proper use is performed according to the operation mode of the operation setting unit 8. For example, in the proper use of the above operation modes depending on the purpose of weighing, there are cases where the weighing object is a truck scale, a crane scale, and a garbage truck. In the case of the truck scale shown in FIG. 12a, the operation mode 1 can be applied because the moving average number adaptation process is completed within a few seconds from when the truck enters the platform, stops, and starts weighing. Operation mode 3 is employed when the crane scale of FIG. In the on-vehicle weighing shown in FIG. 12c, for example, in a garbage truck, weighing is performed for each collection place while traveling around the area. In such a case, if the operation mode 2 is applied and the moving average number adaptation process is completed, the moving average number is updated before the start of measurement at the next collection location. Response delay can be avoided. The operation setting unit 8 sets the operation mode according to the measurement target in this way, and the adaptation processing unit 11 performs an appropriate process.

  In the embodiment described here, the objects to be weighed are the truck scale, the crane scale, and the garbage truck as shown in FIGS. 12a to 12c, but the weighing apparatus 1 of the present invention is not limited to these. When the weight of the object to be weighed 3 is large, it takes time to converge the vibration of the measurement value, so the present invention is more effective. However, when the weight is small, for example, the weighing device 1 used on a table can be used. Therefore, the present invention can be used without any particular limitation on the object of measurement.

DESCRIPTION OF SYMBOLS 1 Weighing device 2 Weighing part 2a Load cell 3 To-be-measured object 4 Support part 5 Amplifier 6 A / D converter 7 Control operation part 8 Operation setting part 9 Display 10 Data holding part 11 Adaptation process part 12 Smoothing process part 13 Storage part

Claims (6)

A weighing unit that outputs an electrical signal corresponding to the weight of the object to be weighed;
A data holding unit for capturing the electrical signal as measurement data;
Smoothing the metric data with a moving average, calculating a moving average number that is a moving average number that minimizes fluctuations in the moving average, and an adaptation processing unit used for the smoothing;
And a display that displays and outputs a value smoothed by the moving average as a measured value. Further comprising an operation setting unit capable of inputting and setting information necessary for calculation in the adaptation processing unit from the outside;
The adaptation processing unit includes:
The moving average number set in advance or input in the operation setting unit is changed to three or more continuous integers including the moving average number in between, and the moving average of the weighing data is calculated by trial,
In each of the moving average number, to calculate a fluctuation evaluation value representing the fluctuation of the moving average,
The weighing device according to claim 1, wherein an adapted moving average number is determined based on each of the fluctuation evaluation values. The moving average is a single moving average or a double moving average;
The said fluctuation | variation evaluation value takes the difference between the data adjacent to the time passage direction with respect to the time series data of the said moving average, and calculates the sum of squares of the said difference. Weighing equipment. The moving average is a double moving average;
The fluctuation evaluation value is obtained by calculating a sum of squares of the differences by taking a difference of data at the same sampling time with respect to time series data of the moving averages of adjacent moving average numbers adjacent to each other. The weighing device according to claim 2. After calculating the adaptation moving average number, the storage unit stores the relation data of the adaptation moving average number and the corresponding measurement value, and further stores the relation data every time the measurement is repeated,
The adaptation processing unit derives a relational expression between the adaptation moving average number and the metric value based on the relation data,
5. When performing a new measurement, the adaptive moving average number is determined using the relational expression according to a mode set in the operation setting unit, and the moving average number is updated. The weighing device according to any one of the above. When the amplitude of the weighing data is not less than a predetermined value and the rate of change of the weighing data is not more than another predetermined value, the weighing data is smoothed by a moving average,
The moving average number of the moving average is changed to a continuous integer of at least 3 including the moving average number in between, and the moving average of each moving average number is calculated by trial,
Calculate a fluctuation evaluation value representing a fluctuation of each moving average,
Determining an adaptive moving average number that is the moving average number that minimizes the variation of the moving average based on each of the fluctuation evaluation values;
A weighing method, wherein a value obtained by smoothing the weighing data by the moving average of the adapted moving average number is used as a weighing value.

Images