JP6965489B2 - Rolling count prediction system and rolling method - Google Patents

Description

The present invention is a technique relating to residential land development, embankment constructed on roads, river embankments, etc., and more specifically, optimal transfer by processing a large amount of data using artificial intelligence AI (Artificial Intelligence). The present invention relates to a rolling compaction number prediction system for predicting the number of compactions and a rolling compaction method using the same.

The required performance (strength, deformation / compression characteristics, water permeability, etc.) of the embankment is set in advance according to the application such as residential land development, roads, river embankments, etc., and the embankment is properly tightened to satisfy this required performance. Hardening is done. Specifically, the embankment material is rolled out by a predetermined thickness (hereinafter referred to as "planned rolling thickness"), and this is referred to as a predetermined number of times by the compaction machine (hereinafter referred to as "planned rolling number of times"). ) Only roll.

Usually, the planned ejection thickness and the planned number of rolling compactions are determined by conducting a test construction in advance. The material (filling material) used in this construction is compacted using the same compaction machine as in this construction, with several types of construction specifications (combination of construction conditions including rolling thickness and number of compactions). By confirming the rolling compaction state by a sand replacement method, an RI (Radio Isotope) measuring method, or the like, the optimum construction specifications (that is, the planned ejection thickness and the planned rolling compaction number) are determined. At this time, the compaction machine is supposed to reciprocate in each of the compaction lanes in which the test yard is divided into a plurality of parts, and therefore, among the construction specifications to be tested, the number of compactions is an even number (2 times, 4 times, 6 times, 8 times). Etc.) is generally planned.

In this construction, the embankment material is sprinkled with the planned sprinkling thickness determined in the test construction, and the compaction is performed at the planned rolling number of times determined in the test construction. When the entire surface is compacted by the planned number of compactions, the compaction work is completed for the time being. Then, the density is measured at the representative points by the sand replacement method or the RI metering method, and if the measurement results at these representative points satisfy the regulations, the construction is finally completed. That is, the actual condition is inspected only at the representative part, and the actual condition is that the construction is completed with the uninspected part left.

Conditions such as compaction energy and soil condition do not completely match between the test construction and the main construction. Therefore, even if the embankment material is rolled out with the planned rolling thickness and the rolling compaction is performed at the planned rolling compaction number, the embankment after the main construction does not satisfy the required performance, that is, the rolling compaction is insufficient, especially in the uninspected part. And there is a risk of over-rolling. In order to solve such problems, new efforts have been made as a quality control method for compacted soil. For example, in Patent Document 1, the acceleration response value of the ground (a value indicating the degree of compaction of the embankment using an acceleration waveform indicating the time change of the vibration acceleration of the vibrating roller) is measured while rolling with the vibrating roller, and the ground is used. The block characteristic value is calculated for each compaction management block based on the relational expression between the acceleration response value and the actual compaction value (test value by sand replacement, etc.), and the quality of the compacted soil is controlled using this block characteristic value as an index. We are proposing the technology to do.

Japanese Unexamined Patent Publication No. 2002-327429

As described above, among the construction specifications tested in the test construction, the number of rolling compactions is planned to be an even number of times, so naturally the planned number of rolling compactions is also determined to be an even number of times. That is, even if the embankment after the main construction satisfies the required performance by rolling compaction a predetermined odd number of times (for example, 5 times), another rolling compaction (6 times in this case) is added. That's why. That is, the construction management method of filling by the planned number of rolling compactions can point out the problem of overrolling by determining the planned number of rolling compactions in an even number of times, and as described above, the conditions of the test construction and the main construction. It is possible to point out the problems of insufficient rolling compaction and overrolling due to the discrepancy between the two.

On the other hand, in the method of Cited Document 1, the quality control of the embankment can be performed in real time and in a face-to-face manner by actually measuring the acceleration response value CCV (Compaction Control Value) of the compacted embankment. However, it is only a method of evaluating the state of the embankment at the time of compaction, and does not indicate how many compactions will be required (or will not be required) in the future. Therefore, if the rolling compaction this time is below the standard even if it is only a little, the next rolling compaction will be performed, and as a result, it may turn into a state of significant overrolling. That is, in the method of Cited Document 1, there may be a case where the compaction cannot be performed the optimum number of times.

The object of the present invention is to solve the problem of the prior art, that is, by predicting the remaining required number of rolling compactions (the number of rolling compactions to be performed in the future), it is possible to roll at the optimum number of rolling compactions. It is to provide a rolling compaction number prediction system capable of providing a high-quality filling by avoiding insufficient rolling compaction and overrolling, and a rolling compaction method using the same.

According to the present invention, artificial intelligence AI is utilized to predict the remaining required number of rolling compactions based on the fluctuating physical property values (the physical properties of the embankment that fluctuate according to the degree of rolling compaction) acquired during the compaction construction. It was made by paying attention to the points, and it is an invention made based on an unprecedented idea.

The rolling compaction number prediction system of the present invention is a system that predicts the required rolling compaction number of the embankment based on the fluctuating physical property value, and is a learning means, a fluctuating physical property value measuring means, a fluctuating physical property value predicting means, and a rolling compaction number calculating means. It is equipped with. Of these, the learning means is a means for learning the learning data set of a large number of sample embankments, and the variable physical property value measuring means measures the variable physical property values of the construction embankment (the embankment to be constructed) and acquires the measured variable physical property values. It is a means to do. Further, the variable physical property value predicting means is a means for predicting the variable physical property value for each number of rolling compactions in the future based on the input data set of the construction filling, and the means for calculating the number of rolling compactions is the reference variable physical property value (target filling is the target). This is a means for obtaining the remaining required number of rolling compactions based on the variable physical property value for each number of rolling compactions predicted in the future. The training data set prepared for each sample embankment includes a plurality of types of physical property values of the sample embankment and variable physical property values for each number of compactions when the sample embankment is compacted, and one of the input data sets is used. Includes multiple types of physical property values of the construction embankment. Then, the learning means generates a fluctuation model showing the relationship between the number of rolling compactions and the fluctuating physical property value based on the learning data set, and the fluctuating physical property value predicting means is used for each input data set, the number of rolling compactions of the construction filling, and the number of rolling compactions. Predict the fluctuating physical characteristics value based on the fluctuating physical characteristics value and the fluctuation model.

In the rolling compaction frequency prediction system of the present invention, the variable physical property value measuring means may acquire the measured variable physical property value for each small area (divided area in which the construction embankment is divided into a plurality of areas). In this case, the fluctuating physical property value predicting means predicts the fluctuating physical property value for each small region, and the rolling compaction frequency calculating means obtains the remaining required number of rolling compactions for each small region.

The compaction number prediction system of the present invention may further include a compaction path setting means. This rolling compaction path setting means is a means for setting a future rolling compaction path based on the remaining required rolling compaction number for each small region obtained by the rolling compaction number calculation means.

The compaction number prediction system of the present invention may further include a compaction machine for compacting the construction embankment. The variable physical property value measuring means in this case is provided in the compaction machine and acquires the measured variable physical characteristic value while moving together with the compaction machine.

The compaction method of the present invention is a method of compacting a construction embankment using the compaction number prediction system of the present invention, and is a method including a compaction step and a variable physical property value measurement step. Of these, in the compaction process, the construction embankment is compacted by the compaction machine until the remaining required number of compactions (calculated by the compaction number calculation means) is reached, and in the variable physical property value measurement process, the construction embankment is compacted by the variable physical property value measurement means. Acquire the measurement fluctuation physical property value of.

The rolling compaction number prediction system and the rolling compaction method of the present invention have the following effects.
(1) Since the compaction can be performed at the optimum number of times, insufficient compaction and overcompacting can be avoided, and as a result, high-quality embankment can be provided.
(2) The rolling compaction may be completed less than the planned number of rolling compactions determined in the test construction. In this case, the construction cost can be reduced, that is, efficient compaction construction can be realized. can.
(3) Since it is possible to provide a reasonable number of rolling compactions regardless of the planned number of rolling compactions based on the conventional reciprocating running, the number of hours required for rolling compaction can be reduced, and as a result, the entire earthwork The cycle time can be shortened.

A model diagram explaining the construction range and small area. The block diagram which shows the main structure of the compaction number prediction system of this invention. The flow chart which shows the main processing flow of the compaction number prediction system of this invention. A model diagram showing "actual rolling number of times-measured variable physical property value" and "future rolling number of times-predicted variable physical property value". A model diagram showing a future rolling path obtained by the rolling path setting means. The flow chart which shows the flow of the main process of the compaction method of this invention.

An example of the rolling compaction number prediction system of the present invention and the embodiment of the rolling compaction method will be described with reference to the drawings.

1. 1. Definitions In explaining an example of an embodiment of the present invention, first, definitions of terms used here will be shown. As described above, one of the features of the present invention is the use of artificial intelligence AI (particularly machine learning). use. Here, for the sake of convenience, the embankment that has been compacted in the past is referred to as the "specimen embankment", the compaction work that is to be carried out from now on is referred to as the "main construction", and the embankment that is the target of this construction. In particular, it is called "construction embankment".

(Physical value of embankment and variable physical property)
The physical property value of the embankment is various attributes related to the embankment, and examples thereof include attributes related to the embankment material and attributes related to the construction. Examples of the attributes of the embankment material include materials (cohesive soil and silt, sand and sandy soil, gravel, etc.) and specifications indicating the state (dry density, water content ratio, porosity, uniformity coefficient, etc.). be able to. On the other hand, as attributes related to construction, the performance and specifications of the compaction machine, the operator of the compaction machine, the number of compactions, the amount of embankment (area and volume), and the construction status (period, number of machines, weather, etc.) are exemplified. be able to.

Among the physical property values of the embankment, those that fluctuate according to the degree of rolling compaction are referred to as "variable physical property values" here. In other words, the fluctuating physical characteristic value is a physical characteristic value that changes each time it is compacted by rolling compaction, and examples thereof include an acceleration response value CCV, a degree of saturation, and a settlement amount. Since the fluctuating physical property value changes each time it is compacted, it is usually acquired by sequentially measuring while compacting. Specifically, if the acceleration response value is CCV, it is acquired by measurement with an accelerometer, etc., the saturation degree is acquired based on the measurement results of a moisture sensor, RI metering method (neutron beam), etc., and the amount of sinking is a satellite observation system. It is acquired by height measurement using GNSS (Global Navigation Satellite System) or total station TS (Total Station). If the variable physical property value is stored (associated) with the acquisition position, the subsequent processing becomes extremely efficient. Therefore, the plane coordinate measurement using the satellite observation system GNSS, total station TS, etc. is performed together with the measurement of the variable physical property value. It is good.

(Measured variable physical property value and standard variable physical property value)
As described above, usually, the fluctuating physical property values are sequentially measured while rolling. Therefore, the fluctuating physical characteristic value obtained by measuring during the main construction is referred to as "measured fluctuating physical property value" here for convenience. Further, considering that the fluctuating physical property value changes each time the compaction is performed, it can be considered that this fluctuating physical property value can be an index for evaluating the construction embankment. That is, a reference variable physical property value is set in advance, and when the measured variable physical property value reaches the standard, it is evaluated that the required performance required by the construction embankment is satisfied. Here, for convenience, the variable physical characteristic value that serves as a reference is referred to as a “measured variable physical characteristic value”.

(Construction range and small area)
Since construction embankments are generally planned over a relatively wide range, it is better to divide the entire embankment and evaluate each part individually, rather than evaluating the embankment quality as a whole. Can be obtained. Here, the plane range of the construction embankment is referred to as "construction range", and each part obtained by dividing the construction range is referred to as "small area". For example, in FIG. 1, a large number (8 × 11 in the figure) of small regions Ms are set by dividing the construction range Cr into so-called meshes. It should be noted that basically, the traveling lane Ln (in the direction of the arrow shown in the figure) of the compaction machine is formed by a plurality of small regions Ms arranged in a row.

2. Rolling Count Prediction System Next, the rolling compaction prediction system of the present invention will be described. FIG. 2 is a block diagram showing a main configuration of the rolling compaction number prediction system 100 of the present invention, and FIG. 3 is a flow diagram showing a main processing flow of the rolling compaction number prediction system 100. In this flow chart, the processing to be performed is shown in the center column, the input information required for the processing is shown in the left column, and the output information generated from the processing is shown in the right column.

FIG. 2 is a block diagram showing the main configuration of the rolling compaction number prediction system 100, and FIG. 3 is a model diagram schematically showing the main configuration of the rolling compaction number prediction system 100. As shown in FIG. 2, the rolling compaction number prediction system 100 includes a learning means 101, a variable physical property value measuring means 102, a variable physical property value predicting means 103, and a rolling compaction number calculating means 104, and further includes learning data set storage. The means 105, the variation model storage means 106, the input data set storage means 107, the measurement variation physical property value storage means 108, the compaction path setting means 109, and the output means 110 such as a display or a printer can be included. It can also be configured to include a compaction machine 111 and a positioning means 112. The rolling compaction number prediction system 100 can be manufactured as a dedicated one, but a general-purpose computer device can also be used. This computer device can be configured by a personal computer (PC), a tablet PC such as an iPad (registered trademark), a PDA (Personal Data Assistance), or the like. Hereinafter, each of the main elements constituting the rolling compaction number prediction system 100 will be described in detail.

(Learning means)
The learning means 101 learns a "learning data set" of a large number of sample fillings, and this learning data set is stored in the learning data set storage means 105. Here, the learning set is a data set prepared for each sample embankment, and the relationship between a plurality of types of physical property values in the sample embankment and the number of rolling compactions and the fluctuating physical property values (hereinafter, "number of rolling compactions-variable physical property values"). It is a data set including.). The number of rolling compactions-variable physical property values is a record when the sample embankment is compacted, and the number of rolling compactions (1 time, 2 times, 3 times, etc.) and the fluctuating physical property values at that time are It represents a relationship.

As shown in FIG. 2, the learning means 101 reads a large number of learning data sets from the learning data set storage means 105 and learns them (for example, deep learning) to obtain a physical property value of the sample filling and the number of rolling compactions. A model for predicting the fluctuating physical property value (hereinafter referred to as “variable physical property value”) is generated using the fluctuating physical property value as a parameter (variable) (Step 10 in FIG. 3). The variation model generated here is stored in the variation model storage means 106 as shown in FIG. As shown in FIG. 3, this variation model should be created before the main construction is carried out.

(Variable physical property value measuring means)
The fluctuating physical property value measuring means 102 measures the fluctuating physical characteristic value of the construction embankment and acquires the measured fluctuating physical characteristic value. As will be described later, the present invention predicts the number of rolling compactions that will be required in the future by comparing the measured fluctuating physical property value with the reference fluctuating physical property value. That is, the measured fluctuating physical property value is acquired while rolling, so it is advisable to set the type of fluctuating physical property value in advance in this construction. For example, when evaluating the fluctuating physical property value as the acceleration response value CCV, an accelerometer or the like is prepared as the fluctuating physical characteristic value measuring means 102, and when evaluating the fluctuating physical characteristic value as the saturation degree, a moisture sensor or the like is used as the fluctuating physical characteristic value measuring means 102. When an RI meter or the like is prepared and the fluctuating physical characteristic value is evaluated as the amount of subsidence, a satellite observation system GNSS or a total station TS is prepared as the fluctuating physical characteristic value measuring means 102. Of course, the evaluation is not limited to one type of fluctuating physical characteristic value, but can also be evaluated by two or more types of fluctuating physical characteristic values (for example, acceleration response value CCV and saturation).

The measured fluctuating physical property value acquired by the fluctuating physical characteristic value measuring means 102 is stored in the measured fluctuating physical property value storage means 108. Since the fluctuating physical property value measuring means 102 acquires the measured fluctuating physical property value each time the number of rolling compactions is repeated, the measured fluctuating physical property value is associated with the number of times of the rolling compaction (hereinafter, referred to as "actual rolling number of times"). Then, it is stored in the measured variable physical property value storage means 108. Specifically, it is stored as a combination of the measured variable physical property value of the first rolling compaction (actual rolling compaction number) and the measured variable physical property value of the second rolling compaction. Further, especially when the construction filling is evaluated for each small area Ms (FIG. 1), the measurement position of the variable physical characteristic value measuring means 102 is positioned by using the positioning means 112 such as the satellite observation system GNSS, and the measurement position is used as the measurement position. It is preferable to associate the measured variable physical property values and store them in the measured variable physical property value storage means 108. Since the measured variable physical property value has the actual number of rolling compactions and the measurement position (planar coordinates), it can be expressed as the measured variable physical property value for each small region Ms (and for each actual rolling number of times).

As shown in FIG. 3, the variable physical characteristic value measurement (Step 30) and the positioning (Step 40) can be performed substantially at the same time (immediately after) as the rolling compaction (Step 20). Can also be done. When measuring and positioning the fluctuating physical property value substantially at the same time as rolling compaction, it is preferable to mount the fluctuating physical characteristic value measuring means 102 and the positioning means 112 on the rolling compaction machine 111. Along with the compaction machine 111 that moves while rolling, the variable physical characteristic value measuring means 102 and the positioning means 112 also measure while moving. On the other hand, when the variable physical property value measurement and the positioning are performed after the compaction is performed, the variable physical characteristic value measurement means 102 and the positioning means 112 are prepared separately from the compaction machine 111, and each of them follows the compaction machine 111. measure.

(Means for predicting variable physical property values)
The fluctuating physical property value predicting means 103 predicts the fluctuating physical characteristic value for each number of rolling compactions in the future. The number of rolling compactions already carried out in this construction is referred to as the "actual number of rolling compactions", whereas the number of rolling compactions to be carried out in this construction in the future is referred to here as the "number of rolling compactions in the future". And. As shown in FIG. 3, when the measured fluctuating physical property value is acquired by the fluctuating physical characteristic value measuring means 102, the fluctuating physical characteristic value predicting means 103 predicts the fluctuating physical property value for each number of rolling compactions in the future (Step 50). However, the specification may be such that the prediction is made after the first rolling compaction and the prediction is made after the second rolling compaction, etc. It may be a specification to predict.

Hereinafter, the process by which the fluctuating physical characteristic value predicting means 103 obtains the predicted fluctuating physical property value for each number of rolling compactions will be described in detail with reference to FIG. First, the variable physical property value predicting means 103 reads out the variable model from the variable model storage means 106, and the relationship between the actual rolling number of times and the measured variable physical property value from the measured variable physical property value storing means 108 (hereinafter, “actual rolling number of times-measured variation”). It is described as "physical characteristic value"), and the "input data set" is read from the input data set storage means 107. Here, the input data set is a data set including a plurality of types of physical property values (attributes related to the filling material used for the main construction and attributes related to the planned construction) in the construction filling, so to speak, prepared before the main construction. It is a dataset that can be used. Therefore, the input data set may be stored in the measurement variation physical property value storage means 108 before the main construction.

The fluctuation model, the actual rolling number of times-measured variable physical property value, and the variable physical property value predicting means 103 that read the input data set use the actual rolling number of times-measured variable physical characteristic value and the input data set as parameters (variables), and the fluctuation model. Is used to obtain the predicted fluctuating physical property value for each number of rolling compactions in the future (hereinafter, referred to as "the number of rolling compactions in the future-predicted fluctuating physical property value"). FIG. 4 is a model diagram showing "actual rolling number of times-measured variable physical property value" and "future rolling number of times-predicted variable physical property value", and "actual rolling number of times-measured variable physical property value" is represented by a solid line. "Future compaction frequency-predicted fluctuation physical property value" is indicated by a broken line. Therefore, in this figure, the number of rolling compactions-predicted fluctuation physical property value is obtained at the end of the third rolling compaction, that is, the actual number of rolling compactions is from the first to the third, and the fourth and subsequent ones are. It is said to be the number of rolling compactions in the future. When the construction embankment is evaluated for each small area Ms (FIG. 1), it is preferable that the variable physical property value predicting means 103 obtains the number of future rolling compactions minus the predicted variable physical property value for each small area Ms.

(Means for calculating the number of rolling compactions)
The rolling compaction number calculation means 104 obtains the remaining required rolling compaction count (how many more rotation pressures should be applied in this construction). As shown in FIG. 3, when the rolling compaction number-predicted variable physical property value is obtained by the variable physical property value predicting means 103 in the future, the rolling compaction number calculating means 104 calculates the remaining required rolling compaction number (Step 60). Specifically, as shown in FIG. 4, the predicted variable physical property value reaches (or exceeds) the standard variable physical property value based on the number of rolling compactions-predicted variable physical property value and the preset standard variable physical property value. The number of rolling compactions (hereinafter referred to as "the number of final rolling compactions") is obtained, and the remaining required number of rolling compactions is calculated from the difference between the current actual number of rolling compactions and the final number of rolling compactions. It is predicted that the predicted variable physical property value will exceed the standard variable physical property value at the 5th (final rolling number of times) rolling, and since the rolling has been performed 3 times (actual rolling number of times) so far, the remaining The required number of rolling compactions is calculated as two times. When the construction filling is evaluated for each small area Ms (FIG. 1), the rolling compaction number calculating means 104 obtains the remaining required number of rolling compactions for each small area Ms. It should be a specification.

As shown in FIG. 3, when the remaining required number of rolling compactions is obtained by the rolling compaction number calculation means 104, the rolling compaction is repeated for that number of times (Step 70), and when the construction embankment is evaluated for each small area Ms, Repeated rolling is performed in all small region Ms until the final number of rolling compactions is reached (Step 80). In this way, by obtaining the required number of rolling compactions (final number of rolling compactions) and performing rational rolling, it is possible to complete an extremely high-quality embankment without causing insufficient rolling compaction or over-rolling. can.

(Rolling path setting means)
As shown in FIG. 1, when the construction embankment is evaluated for each small area Ms, the measured variable physical property value storage means 108 stores the measured variable physical property value for each small area Ms, and the variable physical property value predicting means 103 stores the measured variable physical property value for each small area Ms. It was explained that the number of rolling compactions-predicted fluctuation physical property value is obtained for each rolling compaction, and the rolling compaction number calculating means 104 calculates the remaining required number of rolling compactions for each small region Ms. Therefore, all the small region Ms may have the same required number of rolling compactions, and each small region Ms may have a different required number of rolling compactions. When the required number of rolling compactions is different for each small region Ms, if rolling is performed in the simple straight traveling lane Ln as shown in FIG. 1, the small region Ms that has already reached the final number of rolling compactions may be traveled. Yes, that is, it is possible to exceed the overturning pressure.

The rolling compaction path setting means 109 obtains a future rolling compaction path (route on which the compaction machine 111 travels) based on the final rolling compaction number calculated by the rolling compaction number calculation means 104. Specifically, as shown in FIG. 5, a small region Ms that has not reached the final rolling number of times (hereinafter, referred to as "unfinished small region Ms" for convenience) is extracted and passed through the unfinished small region Ms. Find the optimal (most efficient) rolling path. Examples of the method for obtaining the optimum compaction path include a method for searching the surrounding small region Ms and a method for extracting from all cases.

First, a method for searching the surrounding small area Ms will be described. In this method, starting from the small region Ms where the compaction machine 111 is currently located, one of the unfinished small regions Ms is extracted from the surrounding small regions Ms (8 small regions Ms in FIG. 6). .. Further, after starting from the extracted small region Ms, any one of the unfinished small region Ms is extracted from the surrounding small region Ms. By repeating this, the rolling path is set. By the way, there are cases where a plurality of unfinished small regions Ms are arranged around the starting small region Ms. That is, there are a plurality of options around the small region Ms of the starting point, and a plurality of rolling compaction paths corresponding to these options can be set. In this case, after setting a plurality of rolling compaction paths, it is advisable to select the optimum (for example, the shortest) one from the plurality of rolling compaction paths.

Next, the method of extracting from all cases will be described. In this method, starting from the small region Ms where the compaction machine 111 is currently located, all routes that can be taken from there (regardless of the unfinished small region Ms) are determined. Then, the optimum (for example, the shortest) route is selected from all the obtained routes.

By obtaining the optimum rolling path in this way, the rolling path setting means 109 reduces the small region Ms that causes excessive rolling, as compared with rolling in at least the simple straight traveling lane Ln shown in FIG. This makes it possible to significantly reduce the time required for rolling.

3. 3. Rolling Method Next, the rolling method of the present invention will be described with reference to FIG. The rolling compaction method of the present invention is a method using the rolling compaction number prediction system 100 described so far. Therefore, avoiding explanations that overlap with the contents described in the rolling compaction number prediction system 100, the rolling compaction of the present invention. Only the content specific to the method will be explained. That is, the contents not described here are the same as those described in "2. Rolling number prediction system" including "1. Definition".

FIG. 6 is a flow chart showing the flow of the main steps of the compaction method of the present invention. When this construction is started, while rolling compaction with the positioning means 112 (Step 100), the measured fluctuating physical property value is acquired by the fluctuating physical characteristic value measuring means 102 (Step 200), and the measured fluctuating physical property value is measured by the positioning means 112. Positioning is performed (Step 300). Then, the rolling compaction is repeatedly performed for the remaining required number of rolling compactions determined by the rolling compaction number calculation means 104 of the rolling compaction number prediction system 100 (Step 400), and the rolling compaction is repeatedly rolled until the final rolling compaction number is reached in all the small region Ms. Apply pressure (Step 500).

The compaction frequency prediction system and the compaction method of the present invention can be used for embankment construction, and can be widely used for embankment structures such as roads, river embankments, coastal levees, dams, and embankments. Considering that the invention of the present application provides a high-quality soil structure as a social infrastructure, it can be said that the invention can be used not only industrially but also can be expected to make a great contribution to society.

100 Rolling number prediction system 101 Learning means (of rolling number prediction system) 102 (Rolling number prediction system) Fluctuation physical property value measuring means 103 (Rolling number prediction system) Variable physical property value predicting means 104 (Rolling count prediction system) Rolling count calculation means (of the prediction system) 105 Learning data set storage means (of the rolling count prediction system) 106 (Variation model storage means (of the rolling count prediction system) 107 Input data set storage means (of the rolling count prediction system) 108 Measurement fluctuation physical property value storage means (of rolling compaction number prediction system) 109 Rolling compaction route setting means (of rolling compaction number prediction system) 110 Output means (of rolling compaction number prediction system) 111 (of rolling compaction number prediction system) Pressure machine 112 Positioning means (of rolling compaction number prediction system) Cr Construction range Ms Small area Ln Travel lane

Claims (5)

It is a system that predicts the required number of times of rolling of the embankment based on the fluctuating physical property value which is the physical property value of the embankment that fluctuates according to the degree of rolling.
A learning method for learning a large number of sample embankment learning datasets,
A fluctuating physical property value measuring means for acquiring the measured fluctuating physical property value by measuring the fluctuating physical property value of the construction embankment to be constructed, and
A variable physical property value predicting means for predicting the variable physical property value for each future rolling number of times of the construction embankment based on the input data set of the construction embankment.
The remaining required rolling compaction is based on the reference variable physical characteristic value which is the variable physical characteristic value targeted by the target embankment and the variable physical characteristic value for each future rolling compaction predicted by the variable physical property value predicting means. A means for calculating the number of rolling compactions for obtaining the number of times is provided.
The learning data set prepared for each sample embankment includes a plurality of types of physical property values of the sample embankment and the variable physical property values for each number of times of compaction when the sample embankment is compacted.
The input data set includes a plurality of types of physical property values of the construction embankment.
The learning means generates a fluctuation model showing the relationship between the number of rolling compactions and the fluctuation physical property value based on the training data set.
The fluctuating physical property value predicting means predicts the fluctuating physical property value based on the input data set, the number of rolling compactions of the construction embankment, the measured fluctuating physical property value for each number of rolling compactions of the construction embankment, and the fluctuation model. do,
A rolling compaction number prediction system characterized by this. The variable physical characteristic value measuring means acquires the measured variable physical characteristic value of the small area for each small area obtained by dividing the construction embankment into a plurality of parts.
The variable physical characteristic value predicting means predicts the variable physical characteristic value for each of the small regions.
The rolling compaction number calculation means obtains the remaining required rolling compaction count for each of the small regions.
The rolling compaction number prediction system according to claim 1. Further provided with a rolling compaction path setting means for setting a future rolling compaction path based on the remaining required rolling compaction number for each small region obtained by the rolling compaction number calculation means.
2. The rolling compaction number prediction system according to claim 2. Further equipped with a compaction machine for compacting the construction embankment,
The variable physical property value measuring means is provided in the compaction machine, and acquires the measured variable physical characteristic value as the compaction machine moves.
The rolling compaction number prediction system according to any one of claims 1 to 3, wherein the rolling compaction number prediction system is characterized. A method of rolling the construction embankment using the rolling compaction number prediction system according to any one of claims 1 to 4.
The compaction process of compacting the construction embankment with a compaction machine,
A variable physical property value measuring step for acquiring the measured variable physical characteristic value of the construction embankment by the variable physical characteristic value measuring means is provided.
In the compaction step, the construction embankment is compacted until the remaining required number of compactions determined by the compaction number calculation means is reached.
A compaction method characterized by that.

Images