JPH0489709A - Optimum conveyance control device for raw material yard - Google Patents

Abstract

PURPOSE:To automatize starting and stopping by reasoning and planning a conveyance operation plan for an amount per one shift according to a knowledge object and an operational knowledge of expert operators, stopping automatically when a predetermined conveyance volume has been conveyed completely, outputting the present operation starting command, and at the same time learning an efficiency function about conveyed volumes. CONSTITUTION:According to a knowledge base, a AI tool and a knowledge object stored in files 22-24, an operation means 20 makes a conveyance operation schedule for an amount per one shift by means of a reasoning means 26 so as to store it in a file 25. Then, according to a conveyance operation planning specification inputted from an input means 30, an optimum conveyance control device for a raw material yard monitors an operation progress condition through an operation progress control screen 33. Then, when a conveyed volume has reached a target volume, commands for the finish of the present operation and the starting of the next operation are outputted to a subordinate control device 153 through an operation progress control file 36. Furthermore, an operation on an efficiency function about conveyed volumes is carried out according to a conveyance operation result so that the knowledge object in the file 24 can be renewed.

Description

[Detailed Description of the Invention] [Industrial Field of Application] This invention utilizes a learning-controlled cut-out efficiency function based on the operational knowledge and transportation experience of experienced operators to transfer coking coal temporarily placed in a raw material yard into a blending tank. This invention relates to an optimal conveyance control system for raw materials yards that creates a conveyance work plan for conveying materials to other areas and automatically starts and stops conveyors and loaders based on the results without operator intervention.

[Prior Art] A conventional method of planning a conveyance work plan and a method of automatically starting and stopping a conveyor or a loader will be described below.

First, a method of formulating a transportation work plan will be explained. Here, the transport work plan refers to a transport work schedule for transporting coking coal from the coking coal yard to the blending tank. Generally, a blending tank consists of a collection of 15 to 25 hoppers, and the capacity per tank is 300 to 400 tons. In order to prevent the blending tank from running out of stock, raw coal is transported before the tank is emptied, and the amount transported per operation is 50 to 300 tons.
It's on. In addition, the conveyance efficiency varies depending on the loader's cutting capacity and the shape of the delivery pile, and is 500 to 100.
There is a large variation of 0Ton/hr, and the conveyance time per time is 10 to 30 minutes. There are the following problems in formulating a transportation work plan in such a coking coal yard, and it was only possible to create a transportation work plan for two hours.

(1) Conveyance efficiency varies greatly depending on the cutting capacity of the loader, the amount conveyed, the loader, the brand, and the position of the cut pile (unloading stage), but these factors are taken into account when calculating the conveyance time and planning the work schedule. It was difficult to do. Furthermore, since there is no function to learn a cut-out efficiency function for calculating transport efficiency based on actual work results, there was a large difference in transport time between the planned and actual results.

(2) For this reason, even if we took the time to draw up a work plan for 8 hours, major revisions would have to be made 2 hours later.

Next, a method for automatically starting and stopping the conveyor and loader will be explained.

FIG. 14 is a block diagram showing the configuration of the material yard conveyance control device. In the figure, (150) is the operator interface of the process computer, (151) is the process computer that collects work results and creates work records, and (150) is the process computer that collects work results and creates work records.
52) is the operator interface for the conveyor or loader control device, (153) is the lower control device for sequence control of the conveyor or loader, (154) is the weighing machine for measuring the amount of discharge, and (155) is the raw material The conveyor (15B) is a loader for discharging raw materials.

FIG. 15 is a flowchart showing the operation of the material yard conveyance control device shown in FIG. 14. The explanation will be given below according to this figure.

a. The operator can operate the conveyor (155) or loader (15B).
) operation status via the operator interface (152)
The lower control device (153) is constantly monitored and work settings are made to the lower control device (153) from the operator interface.

b. The operator issues a start-up instruction to the lower control device (158).

The C0 lower control device (153) sequentially starts the conveyor (155) and loader (156).

d. The lower control device (153) constantly monitors the conveyance amount based on the signal (cutting amount) from the weighing machine (154).

e. When the lower control device (153) determines that the transport amount exceeds the target value, it notifies the operator.

f. The operator issues a stop instruction to the lower control device (158).

g. The lower control device (153) sequentially stops the conveyor (155) and the loader (15B). It also transmits the work results to the process computer (153). The work results are displayed on the operator interface (150
) from time to time.

[Problem to be solved by the invention] During transportation work in a raw material yard, the number of tanks (brands)
There are many transport lines, and there is a lot of freedom in selecting the brands and transport lines to be transported. For this reason, transportation work plans had to be developed through trial and error using advanced operational experience (knowledge). Therefore, it takes a huge amount of time to formulate an accurate transport work plan, and it is difficult to do it manually depending on various conditions (in particular, the cutting capacity of the loader, the transport amount, the loader, the brand, and the position of the cutting pile (discharging stage)). It was difficult to create a transportation work plan that could be used effectively in actual operation, taking into account the large differences in cutting efficiency (cutting efficiency), and to create a transportation work plan that lasted one shift (7 to 9 hours). In addition, with the manual planning method, it is impossible to set up and learn a cutting efficiency function based on the actual transportation results, and the accuracy decreases, so it is only possible to draw up a work plan for two hours.

Furthermore, in the past, it was necessary to wait for the completion of the current work and then set the next work one by one, which required the operator to constantly monitor the work status, resulting in very poor work efficiency.

Furthermore, starting and stopping the conveyor and loader all depended on operator operations, and workability was poor in this aspect as well.

Conventionally, due to the above-mentioned problems, it took about 30 minutes to draw up a two-hour transport work plan based on the operator's advanced operating experience. The operator is required to perform work settings one by one in accordance with this transport work plan while constantly monitoring the transport status, and there has been a strong demand for systematization of transport work plan planning and work settings. However, this could not be realized due to the following problems.

(1) The planning of transportation work plans is mainly based on experiential knowledge, and it is difficult to describe this knowledge using conventional programming methods.

(2) There are many combinations of transport brands (tanks) and transport lines, giving too much freedom. For this reason, methods such as linear programming took too long to calculate.

(3) Daily mix plan used to formulate transportation work plan;
The accuracy of yard conditions, perspective planning, cutting efficiency, etc. is not good.

(4) Since planning is done manually, even if conveyors and loaders can be automatically started and stopped without operator intervention, the effect in terms of labor saving is small.

This invention was made in view of this situation,
The advanced operational knowledge of experienced operators regarding transport work plans is converted into a knowledge base, which enables the standardization and transfer of technology, as well as rapid response to operational changes. , different cutting efficiency functions are provided for each cutting pile position (unloading stage), and a function is provided to learn this based on conveyance records to improve planning accuracy.Furthermore, conveyors and loaders are automatically started and stopped. The purpose of the present invention is to provide an optimal conveyance control device for raw material yards that enables the creation of conveyance work plans to be registered as work details and automatically starts and stops conveyors and loaders without operator intervention. do.

[Means for Solving the Problems] The optimum conveyance control device for a raw material yard according to the present invention is capable of adjusting a work plan for conveying coking coal, which has been temporarily placed in piles according to brands in the yard, to a coal blending tank, as well as a daily blending plan. The optimum conveyance control system for raw material yards automatically starts and stops conveyors and loaders without operator intervention, based on changes in blending tank inventory, etc., which change from moment to moment. Based on a knowledge object that includes data such as repair status, perspective cargo handling status, daily blending plan, etc., and a knowledge base that stores each data of this knowledge object and operational knowledge, there is no out-of-stock of blending tanks, and an inference means for formulating a transport work plan so as to maximize the transport efficiency; a means for determining the transport time by providing different cut-out efficiency functions for each transport amount, loader, brand, and delivery position (delivery stage); A learning means that calculates a cutout efficiency function based on work results and updates data on the transport efficiency of knowledge objects, and registers the planning results as work details.
By sending a message to a lower-level system before work begins, conveyors and loaders can be started automatically without operator intervention.
and control means for stopping.

[Function] In this invention, knowledge objects created based on knowledge objects such as daily mixing plans, current yard conditions, perspective plans, conveyor and loader repair plans, cutting efficiency functions, etc., and operations by skilled operators are used. Based on the knowledge base that stores knowledge, a transport work plan for one shift is created for each transport line and registered as the work details.

Meanwhile, the conveyance status is constantly monitored, and when the planned amount of conveyance is completed, lower-level control automatically sends start/stop instruction signals to the conveyor or loader, which are necessary to start the current work, without operator intervention. Output to device. Furthermore, when the conveyance work is completed, the conveyance results are collected by brand and the position of the cutting pile (dispensing stage), and this data can be used to arbitrarily learn a cutting efficiency function.

[Embodiment] FIG. 1 is a block diagram of a transportation work planning device in an optimal transportation control device for a material yard according to an embodiment of the present invention. In the figure, (1o) is an input means for inputting daily mixing plans, repair plans for conveyors and loaders, etc.
20) is a calculation means.

(21) is the I/P buffer of the calculation means (20), (22
) is the file where the knowledge base is stored, (23) is the file where the AI tool is stored, (24) is the file where the knowledge objects etc. described later are stored,
(25) is a file in which a coking coal transport work schedule is stored, and (26) is an inference means. This reasoning means (26) uses the knowledge base of the file (22), the file (
Using the AI tool in 23) and the knowledge objects stored in the file (24), a transport work schedule for one shift (7 to 9 hours) is planned for each transport line, and the work details (brand, transport amount, transport (work start/completion time, etc.) is stored in the file (25).

FIG. 2 is a block diagram showing a transport work management device in the optimum transport control system for a material yard according to the embodiment.

(30) is an input means for registering a transportation work plan as a work specification, and (31) is an operator interface (CI?) for registering a transportation work plan created by an ES (expert system) or manually as a work specification.
T screen).

(32) is an operator interface (CRT screen) for transmitting the tasks registered as work details one by one to the lower control device (153) and for starting and stopping the conveyor (55) and loader (56). (33) is an operator interface (CI?T screen) for constantly monitoring the progress of the transport work. (34) is an ES planning result storage file, which stores the transport work schedule (25) planned by the apparatus shown in FIG. (3
5) is a work details file, in which a plurality of work details are stored for each transport line in the order of transport operations. (36) is a work progress management file, which stores the work details of the current work and the next work for each transport line, and constantly monitors the work status of the current work based on signals from the conveyor and loader control devices. When the amount of conveyance reaches the target amount, an instruction to end the current work and start the next work (starting/stopping the conveyor or loader) is output to the lower control device (153). (40) is a lower control device for sequentially starting and stopping the conveyor and mobile equipment.

The functions of the devices shown in FIGS. 1 and 2 are conceptually shown, and the hardware configuration itself is the same as the conventional device shown in FIG. 14.

Next, the planning of a conveyance work plan, the automatic operation method of the conveyor and loader, and the learning method of the cutting efficiency function will be explained.

FIG. 3 is a flowchart showing the operation of the transport work planning device shown in FIG. First, as a prerequisite, knowledge objects for one shift necessary for planning are created. Knowledge objects (24) are created from daily mixing plan files, yard current status files, parse plan files, conveyor and loader repair plan files, cutting efficiency function files, and the like.

FIG. 4 is an explanatory diagram showing the contents of the knowledge object.

This knowledge object consists of yard current status, daily mix plan, parse plan, conveyor loader repair plan, cutting efficiency function, etc.

Once the knowledge object (24) is created, inference preprocessing (51) is then performed using the knowledge object.

As pre-inference processing, for the purpose of preventing tank stock-outs and improving conveyance efficiency, we perform aggregation calculations for the same brand in the blending tank and calculate the loader operating rate within the planning period.

■Aggregation calculation for the same brand If the same brand is in the tanks, even if one tank is out of stock, the blending will not be stopped. Furthermore, if the same brand is transported continuously, the transport efficiency is better because there is no need for loader movement. For this reason, a plurality of tanks with the same brand are treated as one tank, and calculations for the same brand are performed in aggregation so that a transportation work plan can be formulated.

Example i = 1 Dispense amount - Σ Dispense amount 1 i = 1 1: brand, n: number of cypress with brand ■ Loader operation rate calculation In general, if priority is given to dispensing by a loader with a high operation rate, the conveyance efficiency will be increased. In addition, stock-out of the blending tank is less likely to occur. Therefore, calculate the loader operating rate using the following formula,
Efforts have been made to enable payout using a loader with the highest operating rate possible.

Operation rate of loader j - Σ Blending schedule ff1k / Cutting efficiency function kj: loader, k: brand, m: number of brands dispensed by loader j, where the planned blending amount is the blend by brand to be blended within the planning time The cutting efficiency function is the cutting function (
hl), the cutting peak position (dispensing stage) correction function (h2) in FIG. 6, the stock correction function (h3) in FIG. 7, and the stock correction function (h4) in FIG. Also, hl
is commonly given to all brands as the shape when h2-h3-h4=1, and the time shown in FIG. 5 is determined according to the amount of conveyance. Also h2. H4 is also common to all brands,
The reduction in conveyance efficiency due to the position of the cutting pile (unloading stage) and the amount of yard inventory is corrected. Furthermore, h3 corrects for differences in brands. The meaning of each function will be briefly explained below.

Generally, transport efficiency decreases at the start and end of work.

The cutout function (hl) is a functional representation of this.

In order to reduce the vertical movement of the loader's boom and to reduce fluctuations in the amount of material being delivered, the delivery is divided into 4 to 5 stages from the top of the mountain to the hem. It is done by shaking it in the direction.

In addition, since the width of the hem of the mountain generally increases, the amount of swinging decreases, and the closer to the hem the dispensing efficiency improves. This is expressed as a function by the position (dispensing stage) correction function (
h2). Because the shape of the mountain and the ease of cutting differ depending on the brand, the efficiency of dispensing also varies slightly. The brand correction function (h3) is a functional representation of this. Even when dispensing at the foot of a mountain, dispensing efficiency decreases when the amount of inventory decreases. The yard stock amount correction function (h4) corrects this.

When the inference preprocessing (51) is finished, the knowledge object (2
4) Inference is made using the knowledge base (22) and the inference pre-processing results, and the transportation work schedule (25) for one shift is determined.

FIG. 3 is a flowchart showing a method of formulating each schedule, and the method will be explained below with reference to this diagram.

(1) Inventory amount forecast calculation (52) Raw materials are constantly discharged from the blending tank according to the blending plan. Therefore, the current inventory amount and the inventory amount after a certain period of time are different. Therefore, in order to formulate an accurate transportation work plan, it is necessary to calculate the inventory amount at that time every time one work plan is completed and use it to check the prerequisites for the work plan and check for out-of-stock items. FIG. 9 shows an example of calculation after a certain amount of time has elapsed.

For example, if you start planning at 9:30, the current work (
1-2 at 9:40 when transport to tank l-1) ends.
The stock in the tank is from 9 o'clock stock (333 tons) to the blended amount from 9 o'clock to 9:40 (500 tons/hr * 10
%/100*40m1n/8O-33ton) (300ton).

This value is used as a prerequisite for planning Work Schedule 1.

(2) Calculation of Stock Out Occurrence Time (53) When a stock out occurs, blending cannot be performed and there is no coal to be used in the coke oven, leading to the shutdown of the coke oven. For this reason, the out-of-stock occurrence time is calculated as follows to prevent out-of-stock occurrence.

■The stock-out occurrence time is used as the seed mother (when the same brand is in multiple cypresses, the aggregated calculation is used), and the minimum stock evaluation function (Figure 10) for the blending tank brand determined based on operational knowledge is used. Calculate using

Example: Out-of-stock occurrence time = (Current stock amount - Minimum stock amount) / Mixing amount of brand i Minimum stock Jm = f (Mixing amount of brand i) (Figure 1O) If they overlap, the out-of-stock time is moved forward (for example, the repair start time - 30 minutes is the out-of-stock time).

■Furthermore, when out-of-stock occurs in multiple tanks during a certain time period, the time at which out-of-stock occurs is moved forward.

Example: Advance amount of out-of-stock occurrence time - Out-of-stock occurrence time - 30
min * Number of tanks where stockout occurs. In the above, * means the sign of multiplication, and it is treated in the same way in the following formulas.

(3) Determining the transport line (54) The transport line is determined as follows, taking into consideration the end time of the current work, repair time, etc.

When there are multiple conveyance lines, the conveyance line that finishes its work first is set as the next conveyance line. During repair, treat it in the same way as during work. An example when there are two conveyance lines is shown below.

Here, the solid line represents the current work time, the broken line represents the repair time,
The left end of each line segment represents the start and the right end represents the end.

Example A line 1 line 2 The next conveyance will be line 1.

Example B Line l　　　　　　　　---1-1 Line 2 The next conveyance will be line 2.

(4) Determining the loader to be used (55) The loader to which the brands in the blending tank are delivered differs depending on the position where they are temporarily placed in the yard. For this reason, based on the calculation result of the out-of-stock occurrence time (53) for each loader, the tank brands are arranged in order of earliest out-of-stock occurrence time, and the following evaluation is performed to determine the loader to be used.

(2) If the out-of-stock time is very early, but there are multiple tanks, the loader that can deliver to the tank with the least amount of inventory is used preferentially to prevent out-of-stock.

■; When there are multiple tanks in which there is no stockout but the stockout occurs a little earlier, the loader with the highest operating rate among them is used preferentially to prevent the stockout from occurring.

■; When there is no cypress that falls under ■■, the loader currently in use will be used preferentially, eliminating the time for switching the conveyance line, etc., and improving the conveyance efficiency.

■; When there is no ■■■, a loader with a high operating rate is used preferentially to prevent future stockouts and to improve the transport efficiency of the entire plan.

(5) Determination of tank brand to be transported (56) After determining the transport line and loader, the final tank brand to be transported is determined based on the evaluation function P (x) created based on operational experience.

P (x) = α * k (x) - β * I' (x) α, β: weight constant, X: brand, where k (x), I (r) are inventory determined based on operational experience The holding time evaluation function (Fig. 11) and the switching work evaluation function (Fig. 12) are shown, and the tank brand to be transported is P (x
) is the maximum stock X.

(6) Determination of conveyance amount and time (57) In order to prevent stock-outs from occurring in other tanks during conveyance, first determine a provisional conveyance amount and check for occurrence of stock-outs. The method for determining the transport amount and time will be described below, taking as an example the case where there are two transport lines.

Provisional delivery time - (ditch tank amount - estimated stock amount) / Cutting efficiency operator spare time Here, the estimated stock amount is the brand stock amount in the blending tank, taking into account the amount to be blended until the end of transportation. . Also,
The margin time is a time that takes into consideration (time during which raw materials are not transported (3 to 5 minutes), time for switching loaders and transport lines, etc.) in order to prevent different brands from being mixed into one tank.

In the present invention, a stock-out check is performed based on this provisional delivery time, and the transport amount and time are determined to avoid stock-outs. A specific example is shown below.

■When a stockout occurs in another tank within the tentative transport time, transport amount - cutting efficiency * (time until stockout occurs - margin time) Transport time - time when stockout occurs - transport start time - margin time shall be.

■Another conveyance line is vacant, and 2 tanks are out of stock.
If there are more than one tank, the amount that can be transported before the second tank runs out of stock is taken as the transport amount.

Conveyance amount - Cutting efficiency * (Time until the second tank becomes out of stock - Margin time) Conveyance time - Time when the second tank becomes out of stock - Conveyance start time - Margin time.

■When stockout does not occur, transport is carried out until the trench tank is filled, as follows: transportation amount = trench tank amount - estimated stock amount transportation time = (ditch tank amount - estimated stock amount) / cutting efficiency + margin time.

(7) Planning completion check If the plan for one shift has been created, the results are stored in a file and the planning work is completed. Also, if you are in the middle of planning (1)
Go back and continue your reasoning.

When the planning of one shift's transport work is completed in this way, the planning results are displayed on the work details schedule registration screen (3) according to Figure 2.
1) is registered as a transport work, and thereafter, each time the transport work is completed, one work is registered as a work detail in the lower control device (4).
0) and connect it to automatic operation of conveyors and loaders.

Additionally, there are times when it becomes difficult to continue transport work as planned due to major changes in equipment or operating conditions. As a countermeasure for this, there is a function to re-plan the remaining portion by making use of the transport work results up to now and the mid-way of the transport work plan created so far (a function to re-plan from the middle), and a work detail schedule registration screen. There is a function to correct it.

Next, the learning method will be explained. It varies depending on the conveyance loader, the position of the cutting pile (unloading stage), the brand, the amount of stock in the yard, etc. Therefore, the method for learning the cutting efficiency function is to learn the payout function (hl), the cutting mountain position (payout stage) correction function (h2), the brand correction function (h3), and the yard stock amount correction function (h4). However, in the present invention, the cutting peak position (dispensing stage) correction function (h2) and brand correction function (h3) are learned for the following reasons.

(2) The time shown in Fig. 7 (hl) changes only in proportion to the amount of conveyance and inversely to the cutting efficiency, and the shape of the peaks hardly changes.

■The position of the cutting peak (payout stage) correction function (h2) does not change much, but in order to learn the stock correction function (h3), it is easier to process if this function is also learned.

■The brand correction function (h3) changes over time due to changes in the properties of the brand.

■The yard stock amount correction function (h4) does not have many elements that change.

Figure 13 shows the position (dispensing stage) correction function (h2
) and a method of learning the stock correction function (h3). The following conditions are checked every time the conveyance work is completed, and data is collected only when all conditions are met.

■There is no change in the position of the cutting pile (dispensing stage) during transportation.

■The amount of yard inventory is the cut-off inventory (yard inventory amount correction function)
h4) S value, yard inventory amount when the top inventory is exhausted)
is more than.

■The conveyance amount is sufficiently large.

Then, based on the collected data, learning control is performed for each brand according to the operator's judgment using the following formula.

(a) Learning of the cutting position (dispensing stage) correction function (h2) (learning value) j - (Σ conveyance amount from the 1st stage / cutting function (value in Figure 5)) / n n: 1st stage Number of data (b) Learning of brand correction function (h3) (learning value) b - (Σ (learning value) j) / mm: Number of dispensing stages [Effects of the invention] As described above, according to this invention, conventionally It took about 30 minutes to draw up a 2-hour transportation work plan based on the current situation, the repair status of the conveyor and loader, the perspective cargo handling situation, the daily mix plan, etc., but it was difficult to draw up a 2-hour transportation work plan based on the current situation, the repair status of the conveyor and loader, the perspective cargo handling situation, and the daily mix plan. By organizing the operating knowledge of skilled operators as a knowledge base and providing a cutting efficiency function, it is now possible to accurately plan one shift's transport work in just three minutes. This has led to the standardization and transmission of technology and a reduction in workload. In addition, the created 91"i transport work plan is registered as a work detail on the work detail interval registration screen, the progress of the work is managed on the process computer, and the work details of the next work are sent to the lower-level control device when the work is completed. As a result, the conveyor and loader can now be automatically moved and stopped without operator intervention.Furthermore, a function has been added to learn and control the cutting efficiency function based on the conveyance record, improving the accuracy of conveyance work planning. This led to improvement.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the configuration of a transport work planning device in an optimal transport control system for a raw material yard according to an embodiment of the present invention, and FIG. A block diagram showing the configuration of the work management device, FIG. 3 is a flowchart showing the operation of the transportation work planning device shown in FIG. 1, and FIG. Figures 5, 6, 7, and 8 are explanatory diagrams showing the cutting function (old) used to plan the cutting efficiency, the cutting peak position (paying stage) correction function (h2), and the brand name. Characteristic diagrams of the correction function (h3) and yard inventory amount correction function (h4), Figure 9 is an explanatory diagram showing an example of inventory prediction calculation for blending tank brands, and Figure 10 shows the minimum inventory evaluation function for blending tank brands. Characteristic diagram, Figure 11 is a characteristic diagram of inventory holding time evaluation function, Figure 12
The figure is a characteristic diagram of the switching work evaluation function, and FIG. 13 is a flowchart showing the operation of the learning method. Fig. 14 is a block diagram showing the configuration of a material yard conveyance control device, and Fig. 15 is a block diagram showing the configuration of a conveyance control device for a raw material yard.
It is a flowchart which shows the stop instruction method.

Claims (1)

[Claims] A work plan for transporting coking coal temporarily stored in piles by brand in the yard to a coal blending tank is drawn up based on the daily blending plan and changes in inventory in the blending tank, and the operator In an optimal conveyance control system for a raw material yard that automatically starts and stops conveyors and loaders without any intervention, knowledge objects containing various data such as the current status of the yard, the repair status of conveyors and loaders, the berth cargo handling status, and daily mixing plans are used. , an inference means for formulating a transport work plan so that there is no out-of-stock of the mixing tank and the transport efficiency is maximized based on the knowledge base in which each data of the knowledge object and operational knowledge are stored; and a transport amount. , means to determine the transport time by setting different cutting efficiency functions for each loader, brand, and delivery position, and learning to calculate the cutting efficiency function based on the transport work results and update the transport efficiency data of the knowledge object. and a control means for automatically starting and stopping conveyors and loaders by registering planning results as work details and transmitting them to lower-level systems before starting work. Conveyance control device.