JPH02229391A - Construction work management device for tunnel work - Google Patents

Abstract

PURPOSE:To enable the automatic measurement of working content and working time by comparing the analytic data of characteristics of a working noise in a tunnel collected by a sound collecting means with the standard pattern data characterized by the characteristic working noises accompanied by each working content followed by identification. CONSTITUTION:A working noise is sampled by a microphone 2, and after it is converted into a digital signal through an A/D converter contained in a transmitting device 4, an analytic data is formed. This analytic processing data is inputted to a data analyzing device 10. The data analyzing device 10 is provided with a memory 21, an arithmetic part 22, and a judging part 23, and the difference from the standard pattern data by working content preliminarily stored in the memory 21 is calculated from the analytic processing data obtained in the arithmetic part 22. In the judging part 23, the working content is judged and totalized, based on the calculation result of the arithmetic part 22. The result of judgment and totalization is typed out on a printer 11 at real time and displayed on CRT 12.

Description

DETAILED DESCRIPTION OF THE INVENTION <<Industrial Application Field>> The present invention relates to a device that is capable of managing construction work by collecting and analyzing working sounds inside a tunnel.

<<Background of the Invention>> Generally, in tunnel construction, several of the same unit works are performed successively in a cycle.

The time required for one cycle of this series of unit operations is called cycle time, and in this type of construction, shortening this cycle time will greatly contribute to reducing construction costs.

To do this, it is necessary to accurately measure the cycle time of each task currently being performed, and use this data to identify tasks that can be improved.

Conventionally, the method of understanding this is to station a measurement person at the work site and record the Δ-1 constant by visually observing the work using a clock or stopwatch, or to appoint a person in charge of the work site. In some cases, a button on a recorder must be pressed at the start and end of each task, and data may be compiled using a computer.

Although data processing is often carried out by computers, the measurement of the work itself currently relies on human power.

However, in this type of management method, the following (a) to
There was a drawback mentioned in (c).

[Problems to be solved by the invention] (a) Since measurement relies on human power, errors in button operation and differences in judgment due to individual differences are inevitable, resulting in a lack of accuracy.

(Example) Since measurement personnel are placed at the work site, there is a risk of danger depending on the location where the al-Al is placed.

(c) If a measurement person is dispatched separately from the worker, personnel expenses will be incurred accordingly, causing an increase in costs.

This invention solves the above-mentioned drawbacks, and by collecting and analyzing the sounds generated during various types of work in tunnels, it is possible to automatically measure the work content and working time. The purpose is to provide a construction management device for tunnel construction.

<Means for Solving the Problems> In order to achieve the above-mentioned object, the present invention provides a sound collection means for collecting working sounds generated inside a tunnel, and analyzes the characteristics of the working sounds collected by this sound collecting means. An analysis means as analysis data, a storage means storing standard pattern data that characterizes unique work sounds associated with each work content, and work sounds collected by comparing the analysis data and the standard pattern data. The apparatus includes a work identification means for identifying the work contents to be performed, and a totaling processing means for totalizing the work contents into work types based on the identification result by the work identification means.

Further, in the present invention, the construction management device having the above configuration includes a work confirmation means for confirming that a specific work among the above work contents is being executed, and a work confirmation means for confirming that a specific work among the above work contents is being executed, and a work confirmation means for confirming the work contents when the work contents are identified by the work identification means. A correction means may be added to judge suitability based on the signal of the means and to change or add to the standard pattern data in case of non-conformity.

<<Operation>> According to the above configuration, by analyzing the work sounds generated during the relevant work and comparing the analyzed data with standard pattern data, it is possible to determine what type of work the data indicates. It can be determined whether

By adopting the structure of claim 2, it is possible to check the correct answer rate when the analysis data of the work sound does not fit the standard pattern data, and also to check the standard of the work sound based on the work sound data that generates the corresponding work content. It can provide learning functions such as updating and adding pattern data.

<<Example>> Hereinafter, an example of the present invention will be described in detail using the drawings.

In addition, prior to the example, the construction work of the tunnel was carried out in accordance with the following (1)
) to (6) are repeated work contents, and in each work content, for example, the following heavy machinery is used, and a unique work sound is generated accordingly.

(1) Drilling work (heavy machinery used: drill jumbo, etc.)
Characteristics of working noise...The characteristic sound is the jumbo power supply (fan) sound, and 250Hz is more prominent than others.

When the actual drilling operation begins, the 500, IK, 2KHz sound reaches a level equal to or higher than 250Hz.

(2)Characteristics of the sound of charge loading work...This is the work that is carried out following drilling work.
The operating heavy machinery and features are also almost the same (2501 {Z's excellence). The difference from drilling work is that there is no drilling sound, so the
The level of 0.1K and 2KHz almost never exceeds 250Hz.

(3) Characteristics of blasting work sounds...can be judged by the overwhelming level of sound. There is a difference of 20 to 30 dB or more from other work sounds.

(4) Characteristics of the work sounds during slip-starting and cocking work: Various heavy machinery moves in a variety of ways, so it is difficult to grasp a single characteristic of the work as a whole. If I were to highlight it, it would be the high all-pass level, the fact that various characteristics appear in short cycles, the IK caused by the back buzzer, the severe level fluctuations at 2KHz, and the relative low level at 8KHz. It would be something like that. However, each generated sound has distinct characteristics.

Combi truck: Due to the high engine speed, it has the characteristics of a gentle mountain with a peak of 500Hz when passing.

Wheel loader... working sound is 125Hz and IKHz
stands out as the peak.

Playing power・・・・・・・・・・・・500.1K
, 2KHz has nearly flat characteristics.

(5) Characteristics of the sound of shoring assembly work...Since heavy equipment is the same as drilling and charging, the characteristics are similar.

(6) Characteristics of spraying work sound...The spraying work sound is 4K, 8KH2
It has the distinguishing feature that the level of changes intermittently and increases considerably.

(7) Characteristics of the idle time work noise between (1) to (6): It can be said to be background noise in the tunnel. The 125Hz level, which seems to be due to air conditioning, has an outstanding characteristic, and the All Pass level is 10% higher than during work.
~20dB lower. This is a very stable pattern.

As mentioned above, although there are slight variations in the characteristics of tunnel construction work depending on the heavy machinery used, depending on the content of the work, work noises that are unique to the heavy machinery or the work conditions in which the heavy machinery is used are generated. Therefore, the construction management device of the present invention collects the sounds during the work, analyzes the collected sounds, and compares them with existing standard pattern data.
It determines what kind of work the sound corresponds to, and then categorizes and tallies it.

FIG. 1 shows the overall layout of the construction management device.

In the figure, inside a tunnel shaft 1, there is a microphone 2 (sound collecting means) for collecting working sounds generated inside the tunnel, a receiving device 4 connected to the microphone 2 via a cable 3, and a data transfer device 6. is located.

This data transfer device 6 is connected via a data transfer cable 7 to a data receiving device 9 installed at an external construction management location 8, and the data received by this receiving device 9 is analyzed by a data analysis device 10 consisting of a computer. The content of the work is determined, and the total data is outputted by the printer 11 and simultaneously displayed on the CRT 12.

In addition, a packet transmitter 14 is installed in the LI machine 13 to be inserted into the tunnel pit 1, and the packet transmitter 14 is activated by being linked to the start switch of the heavy machinery 13 or by detecting vibrations during operation. The packet signal is transmitted using radio waves with a unique frequency assigned to the heavy equipment 13 that is being used.

Correspondingly, a packet receiver 15 is disposed inside the tunnel 1, and data received by the receiver 15 is transmitted to the data analysis device via a cable 16, a data transfer device 6, and a data reception device 9. Enter 10.

The construction management device determines the content of the work from sounds collected using a method similar to voice recognition, extracts parameters characteristic of the work from the waveform received by the microphone 2, and performs data analysis in advance. A so-called pattern matching method is employed in which the waveform is compared with standard pattern data stored in a memory provided in the device 10 to find out which waveform it is similar to.

More specifically, as shown in FIG. 2, the working sound is sampled by the microphone 2 and converted into a digital signal via the A/D converter 20 built into the receiving device 4.

This digital signal has a sampling period of 50, for example.
μS. 1,024 points, and the quantization level is ΔL-
Encode with 1/4096, that is, 12 bits.

The receiving device 4 performs preprocessing such as time window processing and Takagi emphasis on this signal, obtains a power spectrum, and performs Kebs drum processing to create analytical data.

This analyzed data is input to the data analysis device 10. The data analysis device 10 includes a memory 21, a calculation section 22, and a determination section 23. Then, the calculation unit 22 calculates the difference between the obtained analysis processing data and the standard pattern data (12 bits, 512 points) for each work content stored in the memory 21 in advance, and calculates the obtained numerical value. It is cut at a certain threshold and the value exceeding this threshold is marked as “1”.
"The judgment section 23 integrates the frequency.
The work content is determined based on the calculation results in step 2, and the results are totaled. Then, the results of this judgment and totalization are typed out in real time on the aforementioned printer 11 and displayed on the CRT 12.

FIG. 3 shows the specific contents used for the determination, and the work contents (1) to (7) are stored in the memory 21.
) A plurality of standard pattern data l(A-A') to ．It extracts sound pressure levels by frequency.

Then, in the calculation section 22, the difference between the analysis processing data and each standard pattern data is calculated.

This result is then cut at a preset threshold,
Values exceeding this value are set to "1" and the first counter ((A-
An) to Xn).

This first counter has a time limit, and each work type (
Second counters A-X provided corresponding to each of 1) to (7)
The total value is transferred to and calculated as a debt.

In the determination unit 23, a unique determination function is programmed according to the frequency and distribution of the counters A-X, and this function specifies the work content according to the frequency and distribution of the integrated values accumulated in each counter A-X. The work contents are aggregated and output to the printer 11 as well as displayed on the CRT 12.

In the above processing, consistency between the analysis processing data and the standard pattern data becomes a problem.

In other words, if the sampled analysis data and standard pattern data match well, the work content will be analyzed accurately.

However, in actual work, the sound at a work site changes in a complex manner due to the presence of heavy machinery, etc., so it is difficult to create standard pattern data for each work site, and the consistency of the data created is difficult. Difficult to confirm.

Therefore, the data analysis device 10 is programmed with a learning function and executes the following processing.

First, by receiving the transmission status of the packet transmitter 14 attached to the heavy equipment 13, it is confirmed that the work sound currently being collected is generated from the corresponding heavy equipment 13. Therefore, if the work content determined by the determination unit 23 based on the analysis data and the work content confirmed by the packet transmitter 1 match, the standard pattern data stored in the memory 21 is unchanged. It will be possible to use it.

On the other hand, if the work content determined by the determination unit 23 and the work content confirmed by the packet transmitter 14 do not match, the standard pattern data in the memory 21 cannot be used as is. Either the content is changed or the determination function in the determination unit 23 is changed.

Specific corrections to the standard pattern data can be made, for example, in drilling work of type (1) mentioned above, when the type of drill used, the characteristics of the face change, or the drills converge with each other. The working sound changes depending on the memory 11.
The standard pattern data (A-A') corresponding to the drilling operation stored in the drill must also be changed in response to changes in the drill or face.

However, since the signal from the packet transmitter 14 confirms that the corresponding work type is actually being performed,
It is also possible to erase the standard pattern data previously stored in the corresponding areas A to A1 of the memory 21, extract its features from the actually obtained analysis data, and write them as new standard pattern data.

As a method for creating standard pattern data at this time, first, the work identification data from the packet transmitter 14 and the packet receiver 15 is positioned as absolute information (correct information), and the transition points of the work are clarified.

Then, the transitions (points of change) in the work are classified as work content in the same category, that is, if it is spraying, it is classified as spraying work. Work sound information actually collected by category is stored in a memory 21.

The storage format at this time is, for example, by setting a sampling time of around 1 second, and summing up the work sounds obtained within this period into cumulative frequencies of sound pressure levels by frequency.
The frequencies at 0% and 95% are Ls, Lso.
Store as L95.

Also, for the work sounds obtained within the above sampling time, the cumulative frequency at the sound pressure level without specifying the frequency is totaled, and Ls. The frequencies of Lso and L95 are stored, and such data is collected for each work content in one cycle.

The data created in this way is used as standard pattern data, and is replaced with new data by transferring it to the above-mentioned A-A'. At this time, in order to delete the same type of new data as the old data, a comparison operation is performed and processed. Then replace it.

In addition, although there is a large number of standard pattern data in the same category, the standard pattern data that more frequently exceeds the threshold when correctly answering the work judgment increases the weight of the judgment function, and the data that has little involvement in the judgment It automatically deletes the data and moves those with a high probability of misjudgment into a predetermined memory (memory that is completely unrelated to calculations).

It should be noted that data resulting in an erroneous determination is either re-entered into the AA' memory after a human selects the category, or is deleted.

As described above, by adding a learning function (correction function) such as changing or adding memory contents or changing the judgment function, the accuracy rate corresponding to the corresponding task increases and the practicality of the measuring device is increased. This means that you can do it.

In addition, in the above embodiment, as a method of creating standard pattern data, pattern data was used in which features were extracted for each work, but a pattern with partial emphasis by dividing it by frequency and several types of shapes were intentionally set. A combination of patterns created in this manner may be used as standard pattern data. In this case, the values exceeding the threshold value are counted for each standard pattern, and the type of work can be determined from the distribution pattern of this cumulative amount.

Additionally, the ability to check the actual work being done using radio waves and make corrections based on that information is especially necessary in the initial state of the work site to be analyzed, but once the work has started and each time the cycle is repeated, Since the accuracy rate increases, it is also possible to use radio wave confirmation only for the necessary period and then remove it.

Furthermore, confirmation of the work actually being performed may be performed not only by packet communication but also by other means.

<<Effects of the Invention>> As explained in detail using the embodiments above, the tunnel construction construction management device according to the present invention analyzes the work sounds generated during the relevant work, and combines the analyzed data with standard pattern data. By comparing the data, it is possible to automatically determine what kind of work the data represents.

Therefore, in this invention, compared to the conventional method of manually measuring the type of work, it is possible to obtain more accurate data, and there is no danger as it is only necessary to install a sound collection means at the work site. , monitoring costs can be reduced, the content of unit work can be measured in detail according to the measurement accuracy of the sound collection means, and furthermore, the content of work currently being performed at a remote location can be immediately grasped. There are many advantages.

In addition, in this invention, by adopting the structure of claim 2, it is possible to confirm the accuracy rate in cases such as when the analysis data of the work sound does not fit the standard pattern data, and also to Updated standard pattern data for work sounds based on sounds. It can provide learning functions such as adding information, and can accurately grasp the work situation depending on the site.

[Brief explanation of the drawing]

Figure 1 is an explanatory diagram showing the arrangement of measuring equipment used for construction management of tunnel construction, Figure 2 is a block diagram showing the system configuration of the measuring equipment, and Figure 3 shows the specific details of the determination of work sound data. FIG. l... Tunnel pit 2... Microphone (sound collecting means) 4... Receiving device if (analysis means) 10... Data analysis device (identification means) 13... Heavy equipment 14... Packet transmitter 15...Packet receiver

Claims (2)

[Claims] (1) A sound collection means for collecting the working sounds generated inside the tunnel, an analysis means for analyzing the characteristics of the working sounds collected by this sound collecting means and generating analytical data, and unique working sounds accompanying each work content. storage means for storing standard pattern data characterized by; a work identification means for comparing the analysis data and the standard pattern data to identify the work contents corresponding to the collected work sounds; and identification by the work identification means. 1. A construction management device for tunnel construction, comprising a totaling processing means for totalizing work contents into work types based on results. (2) The construction management device according to claim 1 further comprises a work confirmation means for confirming that a specific work among the work contents is being executed, and a work confirmation means when the work content is identified by the work identification means. 1. A construction management device for tunnel construction, comprising: a correction means for determining suitability based on a signal of the standard pattern data, and for changing/adding the standard pattern data in the case of non-conformity.