JP2014017760A - Analog input device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an inexpensive analog input device that suppresses a dead time at a change in moving average frequency and reduces a variation between moving average values prior to and subsequent to the change in moving average frequency.SOLUTION: In the analog input device having AD conversion means for converting an analog signal input at a constant sampling period to digital data to be output as an A/D conversion value, storage means for saving the A/D conversion value, and arithmetic processing means for executing a moving average process of such A/D conversion values saved in the storage means at a set moving average frequency to calculate a moving average value, the arithmetic processing means calculates moving average values at two or more different moving average frequencies, respectively, and selects one of the moving average values thus calculated to be output in accordance with a difference between the moving average values thus calculated and trend data on moving average values calculated at a set one of the moving average frequencies.

Description

  The present invention relates to an analog input device used for a programmable controller or the like, and in particular, it has excellent followability at the time of rising and falling of an input signal, while measuring a stable value by suppressing blurring due to disturbance of an input value in a steady state. It is related with the analog input device which can do.

  Conventionally, a programmable controller inputs an analog signal from a device to be measured and grasps its state, or inputs an analog signal indicating the operation state of the device from a device to be controlled, and a control signal based on this input signal. Is generated and output to the controlled device.

  FIG. 8 is a schematic configuration diagram of a programmable controller 10 including an analog input module (analog input device) 1.

  In general, the programmable controller 10 performs input / output with an interface specification determined between the CPU module 3 that executes arithmetic processing according to an application, the analog input module 1 that inputs an analog signal from an external device, and other external devices. Modules such as an output interface (I / O) module 4 are provided, and these modules exchange data via the bus 2.

  First, the CPU module 3 sets operation parameters (hereinafter simply referred to as “parameters”) in the bus interface memory (bidirectionally readable / writable memory) of the analog input module 1 when the system is activated. This parameter specifies the operation content of the analog input module 1, and the analog input module 1 passes an analog signal input based on this parameter to the CPU module 3.

Hereinafter, the configuration and operation of the analog input module 1 will be described.
An analog signal from an external device is taken in via an input circuit 16 from an input terminal 17 corresponding to each channel of the analog input module 1, and one analog signal selected by the multiplexer 15 is digitalized by the A / D converter 14. The signal is converted into a signal and input to an MPU (arithmetic processing means) 12 that executes arithmetic processing. The MPU 12 stores the A / D conversion value in a RAM (storage means) 13b and performs arithmetic processing such as filter processing and scale conversion processing on the A / D conversion value according to the parameters set by the CPU module 3. After that, the data is written into the bus interface memory 11.

  The MPU 12 executes the above processing at a constant input period (sampling period) for the analog signals of all channels while switching the multiplexer 15.

  As for the filter processing of the analog input module 1, conventionally, moving average processing is performed in order to effectively reduce the influence of disturbance superimposed on the input signal. (For example, see Patent Document 1).

  In order to perform this moving average processing, the analog input module 1 forms a ring buffer in the RAM 13b for each channel. This ring buffer is a storage area for analog input values that are address-mapped in the RAM 13b. As shown in FIG. 9, the ring buffer has an address space for the maximum number of moving averages (REC_MAC) that can be set. It is stored in the storage area of the next address. When the address space allocated to the ring buffer is stored up to the last address, the first address (datarec) is overwritten in the return order.

  The RAM 13b stores an area (sum) for storing the total value of A / D conversion values corresponding to the number of moving averages for each channel, and a moving average value obtained by dividing the total value by the number of moving averages. It also has an ave.

  Hereinafter, the processing contents of the initialization function, data input function, and moving average processing function of the MPU 12 will be described with reference to FIG.

  FIG. 10 shows the processing procedure for one channel, but for the analog input module 1 having a plurality of channels, the same processing is executed for the other channels.

  The MPU 12 executes an initialization process when activated by powering on the system, a reset process, or the like. As an initialization process, the ring buffer, the total value storage area sum, and the moving average value storage area ave are cleared to zero. Further, the value datarec of the head address of the ring buffer is set to the address pointer newrecptr indicating the area in which the latest A / D conversion value is written in the ring buffer (S101).

  When the timing (sampling period) for input processing of the analog signal arrives (S102), the A / D conversion value output from the A / D converter 14 is acquired, and the A / D conversion value is indicated by the address pointer newrecptr. Store in the address (S103).

  Next, it is determined whether or not the moving average count has been changed (S104). If changed, the total value storage area sum and the moving average value storage area ave are cleared to 0 (S105).

  Thereafter, a value obtained by subtracting the moving average number REC_N from the address pointer newrecptr is set in the address pointer oldptr (S106). At this time, if the address pointer oldptr becomes smaller than the head address datarec, the ring buffer size REC_MAX is added to the address pointer oldptr (S107, S108).

  Then, the value stored in the address pointer oldptr is subtracted from the value stored in the total value storage area sum (S109), and the value stored in the address pointer newrecptr is added to this (S110). Thereafter, a result obtained by dividing the value stored in the total value storage area sum by the moving average number REC_N is stored in the moving average value storage area ave (S111). By this process, a moving average value including the newly sampled A / D conversion value is calculated.

  Next, the address pointer newrecptr is incremented (S112). At this time, if the address pointer newrecptr exceeds the final address of the ring buffer (S113), the head address datarec of the ring buffer is set to the address pointer newrecptr (S114). Thereafter, the process proceeds to step S102, and the next sampling period is awaited.

  In the above process, when the moving average value storage area ave is stored in the RAM 13b, the value stored in the area is written in the bus interface memory 11 and can be read from the CPU module 3. Of course, the moving average value storage area ave may be stored in the bus interface memory 11.

  FIG. 11 shows changes in values of a ring buffer having 12 A / D conversion value storage areas, a total value storage area sum, and a moving average value storage area ave for the case of three moving averages. In FIG. 11, (1) to (12) indicate the sampling order. Further, the shaded portion of the ring buffer indicates an area that is a calculation target of the moving average value.

Japanese Patent Laid-Open No. 7-110652

  In general, analog measurement has good responsiveness to changes in the rise and fall of the analog input, while there is a demand to measure a stable value while suppressing blurring of the input value due to disturbance (noise) and other factors. .

  However, as shown in FIG. 12, when the moving average number is small, the moving average has good response at the time of rising / falling, but since the filter effect is weak, the fluctuation of the input value due to disturbance or the like becomes large. Conversely, when the number of moving averages is large, the fluctuation of the input value becomes small, but the responsiveness deteriorates. As described above, since the responsiveness and the fluctuation of the input value conflict, it is necessary to use the moving average number of times depending on the condition.

  On the other hand, in the conventional moving average process as described above, if the moving average number is changed by a command from the CPU module 3, the value stored in the total value storage area sum at that time cannot be used. A process of clearing the area sum and the moving average value storage area ave to 0 is performed. For this reason, when the number of moving averages is changed in the middle, the moving average value once becomes 0 as shown in FIG. Therefore, the continuity of the input values cannot be ensured at the parameter switching timing, and the moving average value cannot be used for a certain time from the time of change in the CPU module. For this reason, there is a problem that dead time occurs.

  In order to cope with this problem, for example, a method of inputting one analog output from an external device to two channels, calculating with different moving average times, and switching a moving average value to be used as appropriate can be considered. However, this method increases the cost because the number of channels increases. In addition, in order to receive 4-20mA output as an analog signal, the common input type cannot be input to multiple channels, so an isolated channel-to-channel analog input method must be adopted, which also causes cost increases. become.

  The present invention has been made in response to such a conventional situation, and provides an inexpensive analog input device that suppresses a dead time when the moving average number is changed and has a small fluctuation of the moving average value before and after the moving average number is changed. For the purpose.

  To achieve the above object, the analog input device of the present invention converts an analog signal input at a constant sampling period into digital data and outputs it as an A / D conversion value, and the A / D conversion. Analog input having storage means for storing a value, and arithmetic processing means for calculating a moving average value by executing a moving average process of A / D conversion values stored in the storage means for a set number of moving averages The calculation processing means calculates a moving average value by at least two different moving average times, and calculates a difference between the calculated moving average values and a set moving average number of times. One moving average value is selected from the calculated moving average values according to the trend data of the average value, and is output.

  Here, the trend data means data that can grasp the change tendency, and includes the change amount or change rate of the current value with respect to the previous value.

  Preferably, two threshold values (first threshold value and second threshold value) for determination can be set from an application program or an external device, and the arithmetic processing means determines that the difference between the calculated moving average values is the first threshold value. Or the trend data exceeds a second threshold value, the moving average value calculated by the smaller moving average number is output, and the difference between the calculated moving average values and the trend data are When both do not exceed the respective threshold values, it is preferable to output the moving average value calculated by the larger moving average number.

  In the present invention, a moving average value of A / D conversion values corresponding to the moving average number is calculated for each moving average number. And the continuity of the data at the time of switching is ensured by switching the moving average frequency | count used as an output by the difference of the respectively calculated moving average value. Furthermore, by using the trend data of the moving average value as the switching condition of the moving average count, the analog input signal rises and falls quickly without increasing the number of channels, and after reaching the steady state (saturated state) Can increase the filter effect by increasing the number of moving averages.

  According to the present invention, it is possible to suppress the blurring of the moving average value before and after changing the moving average number by properly using the moving average number (filter strength) by software processing, thereby preventing dead time when the moving average number is changed. And a highly reliable system can be constructed at low cost.

  Also, excellent response and noise resistance can be realized by reducing the moving average number in a region where the change of the analog signal is large and increasing the moving average number in a region where the change of the analog signal is small. Furthermore, it is possible to ensure the continuity of input values by calculating the moving average value with two different moving average times and switching the moving average value when the difference between both moving average values is below a certain value. it can.

It is explanatory drawing of the ring buffer other data storage area by embodiment of this invention. It is a figure which shows the data structural example of the operation parameter by embodiment of this invention. It is explanatory drawing of the determination table used by the moving average process by embodiment of this invention. It is a flowchart which shows the moving average process procedure by embodiment of this invention. It is explanatory drawing which shows the continuity of the moving average value by the result of the moving average process of FIG. It is explanatory drawing of the operation parameter by the application example 1 of embodiment of this invention. It is explanatory drawing of the operation parameter by the application example 2 of embodiment of this invention, Fig.7 (a) is explanatory drawing of the input range used as a dead zone, FIG.7 (b) is for implement | achieving a dead zone process It is a figure which shows the structure of an operation parameter. 1 is a schematic configuration diagram of a programmable controller 10 including an analog input module (analog input device) 1 according to the prior art and the present invention. It is explanatory drawing of the data storage area | region other than the ring buffer used for the conventional moving average process. It is a flowchart which shows the conventional moving average process sequence. It is a figure which shows transition of the data of each data storage area in the case of a moving average of 3 times, and a moving average value. It is explanatory drawing of the difference in the transition of a moving average value in case a moving average frequency differs. It is explanatory drawing of the discontinuity of the moving average value at the time of the moving average frequency setting value change in the conventional moving average process.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. The embodiments described below are preferred specific examples of the analog input device of the present invention, and may have various technically preferable limitations. However, the technical scope of the present invention particularly covers the present invention. As long as there is no description which limits, it is not limited to these aspects. In addition, the constituent elements in the embodiments shown below can be appropriately replaced with existing constituent elements and the like, and various variations including combinations with other existing constituent elements are possible. Therefore, the description of the embodiment described below does not limit the contents of the invention described in the claims.

An analog input module (analog input device) 1 according to the present embodiment is realized by the hardware configuration shown in FIG. Compared to the prior art, the configuration of the ring buffer, the parameters set from the CPU module, and the moving average processing procedure are different. Hereinafter, the same elements are denoted by the same reference numerals, description thereof is omitted, and differences from the prior art will be mainly described.
(Ring buffer configuration)

  FIG. 1 shows the configuration of a ring buffer and other data storage areas according to this embodiment. Conventionally, only the total value of A / D conversion values for the number of moving averages set as a parameter and the average value calculated by dividing this total value by the number of moving averages are stored, but in this embodiment, The total value of the A / D conversion values and the moving average value for all the moving average times that can be set from the CPU module can be stored.

FIG. 1 shows an example in which the total value and the moving average value are stored for the moving average times 2 to 256, respectively, but the moving average number is not limited to this range. Further, when the number of moving averages increases and the load of computer processing increases, the number of moving averages that can be set from the CPU module 3 is limited, or the total is only calculated for a plurality of moving averages that are actually parameterized. By calculating the value and the moving average value, the processing load can be reduced and the memory space can be saved.
(Parameter)

  The parameters according to this embodiment are shown in FIG. The difference between the moving average value by the first moving average number (B), the second moving average number (A), the moving average value by the first moving average number (B) and the moving average value by the second moving average number (A) Parameters such as a threshold value (threshold value 1), a threshold value of the moving average value change amount by the second moving average number (B) (threshold value 2), and whether or not a moving average setting combination can be made from the CPU module 3 at the start-up To the bus interface memory 11.

  Although the procedure of the moving average process will be described later, these parameters are generally used in the moving average process as follows.

  The moving average value based on the first moving average number (B) is defined as a moving average value (B), and the moving average value based on the second moving average number (A) is defined as a moving average value (A) (where B <A). In this case, the rising / falling change amount of the analog signal is grasped from the difference (ΔB) between the current value and the previous value of the moving average value (B).

  Further, it is determined whether or not the analog signal has reached a steady state (saturated state) based on a difference (ΔA) between the moving average value (B) and the moving average value (A).

  By determining where in the determination table of FIG. 3 the combination of the difference (ΔB) and the difference (ΔA) corresponds, one of the moving average values is selected and passed to the CPU module 3. Specifically, if both the difference (ΔB) and the difference (ΔA) are below the threshold values, it is determined that the steady state has been entered, and the moving average value (A) having the larger moving average count is determined as the CPU module 3. To pass. In the case of other conditions, it is determined that the analog signal is changing between rising and falling, and the moving average value (B) having the smaller moving average count is passed to the CPU module 3. Note that if the two moving average times are not switched, “average setting combination availability” is set to “no”.

The parameters of the analog input module 1 include parameters related to scale conversion processing and alarm processing in addition to the above, but these are not the objects of the present invention, and will not be described.
(Moving average processing)

Next, the procedure of the moving average process according to the present embodiment will be described with reference to FIG.
First, the initialization process clears the ring buffer, the total value storage area sum, and the moving average value storage area ave, as in the conventional FIG. However, since sum and ave are secured for a plurality of moving average times as shown in FIG. 1, all of these are cleared to zero. In addition, the head address value datarec is set in the address pointer newrecptr (S201).

  The subsequent steps S202 to S213 correspond to the steps S102 to S114 in FIG. However, the total value calculation process and the moving average value calculation process shown in steps S205 to S210 in FIG. 4 correspond to steps S106 to S111 in FIG. 10, respectively, but in this embodiment, the moving average number REC_N is a variable. These processes are repeated for all the moving average times (S204a, S204b). Incidentally, the increment of the address pointer newrecptr and the cyclic processing (S211 to S213) in the ring buffer are the same as those in FIG.

  The moving average calculation process according to the present embodiment includes a process of automatically selecting a moving average value based on a more appropriate moving average number out of the two moving average numbers (A, B) after the above process. It is characterized by that.

  Specifically, after steps S212 and S213, the difference (ΔB) between the current value and the previous value of the moving average value (B) based on the smaller moving average number of the two moving average times set as parameters is calculated. In addition to the calculation (S214), the difference (ΔA) between the moving average value (B) and the moving average value (A) based on a larger average number is calculated (S215).

  When at least one of the difference (ΔB) and the difference (ΔA) exceeds the threshold 2 or the threshold 1, the moving average value (B) is written in the bus interface memory 11 (S218), and the difference ( If neither ΔB) nor the difference (ΔA) exceeds the threshold value 2 or the threshold value 1, the moving average value (A) is written in the bus interface memory 11 (S219).

  As shown in FIG. 5, when the change in the analog input signal is gentle, the processing of the above steps S214 to S219 selects the moving average value (A), and when the change in the analog input signal is steep, The moving average value (B) is selected. In the above processing, the moving average value (B) change amount is used for selecting the moving average value. However, the change rate is used instead of the change amount, and the movement is performed when the change rate exceeds the threshold. The average value (B) may be selected.

  According to the present embodiment, the blur of the moving average value at the time of switching of the moving average number can be reduced (below the threshold value 1), and the CPU module 3 depends on whether or not it is the switching timing of the moving average number. Instead, the moving average value obtained from the analog input module 1 can always be used.

  The CPU module only needs to set two types of moving averages as parameters in advance according to the system, and it is not necessary to monitor the trend of the input signal and switch the parameters, reducing the load on the CPU module. can do. Of the two types of moving averages set as parameters, the larger moving average number is set so that the number of times × sampling period is one period or more of the transient response of the input signal. Therefore, it is possible to prevent the moving average number from being frequently switched.

Further, the moving average number for calculating the amount of change or the rate of change (ΔB) and the moving average number for calculating the difference (ΔA) between the moving average number before and after the switching do not necessarily match. For example, ΔB is The moving average value is calculated using the moving average value of the moving average number of times 3, and ΔA is calculated using the moving average value of the moving average number of times of 16 times and the moving average number of times of 128 times, and the moving average value passed to the CPU module 3 is The moving average value may be either the moving average number of times 16 or the moving average number 128.
(Application 1)

  In the above-described embodiment, the two moving average times are automatically switched depending on the rate of change and the amount of change. However, a larger number of moving average times may be switched using the type of control or the like. .

  As shown in FIG. 6, for example, when the rate of change is in a relatively large range (above 10% to 50% or less), even if the moving average number is switched, the moving average number of relatively small values (16 times). , And use this to perform moving average processing. If the rate of change is in a relatively small range (10% or less), select a relatively large number of moving averages (128 times) and use this to move Perform averaging.

In this way, the moving average number to be selected is registered in association with the change rate of the moving average number, the moving average number corresponding to the change rate is extracted, and the moving average process is performed using this. Thus, a more adaptive filter function can be realized.
(Application example 2)

  Further, as shown in FIGS. 7A and 7B, two types of setting values may be added to the parameters as dead zones.

  Specifically, as shown in FIG. 7A, a predetermined range (a1 to a2, b1 to b2) immediately before the analog input signal during rising / falling reaches a steady state (saturated state) is set as a dead zone region. Set (first dead zone set value). For example, when a control output feedback signal is input, a certain range including the target value can be set as the dead zone.

  That is, the first dead zone setting value (a1 to a2 or b1 to b2) is a preset threshold range. The second dead zone set value is a switching prohibition time. When the moving average value (B) or A / D conversion value based on the smaller moving average number is entered in the threshold range specified by the first dead band setting value, the movement is performed for a certain period of time (or a certain number of samplings). Only the transition to the larger average number is possible, and the transition from the larger moving average number to the smaller one is prohibited. As a result, unnecessary switching of the moving average frequency during a transient response can be prevented and stable system operation can be realized.

1 Analog input module (analog input device)
2 Bus 3 CPU module 4 I / O module 10 Programmable controller 11 Bus interface memory (storage means)
12 MPU (arithmetic processing means)
13a ROM
13b RAM (storage means)
14 A / D converter (A / D conversion means)
15 Multiplexer 16 Input circuit 17 Input terminal

Claims (4)

AD conversion means for converting an analog signal input at a constant sampling period into digital data and outputting it as an A / D conversion value;
Storage means for storing the A / D conversion value;
Arithmetic processing means for calculating a moving average value by executing a moving average process of A / D conversion values stored in the storage means for a set number of moving averages;
An analog input device having
The arithmetic processing means calculates a moving average value by at least two different moving average times, the difference between the calculated moving average values, and the trend data of the moving average value calculated by the set one moving average number To select and output one moving average value among the calculated moving average values.   When the difference between the calculated moving average values exceeds a first threshold value, or the trend data exceeds a second threshold value, the arithmetic processing means calculates the moving number calculated by the smaller moving average number of times. When the average value is output and the difference between the calculated moving average values and the trend data do not exceed the respective threshold values, the moving average value calculated by the larger moving average number is output. The analog input device according to claim 1.   The analog input device according to claim 1, wherein the arithmetic processing unit selects a moving average number used for moving average processing based on the trend data.   The arithmetic processing means determines whether the A / D conversion value or the moving average value to be output is within a set range, and if it is within the set range, the moving average to be output for a certain period of time. The analog input device according to any one of claims 1 to 3, wherein a change in the number of moving averages used for calculating a value is prohibited.

Images