Method for automatically controlling a container transport device with one or more conveyor belt for adjusting container

Abstract

The invention relates to a method for automatically controlling a container transport device having one or more conveyor belts (17-28) in order to adjust the container density of containers transported on the one or more conveyor belts. The method comprises detecting distance data of the containers on the one or more conveyor belts at different times in each case by means of a distance measuring device, transmitting the distance data to a control device by means of the distance measuring device, processing the distance data into control data in the control device, and non-linearly controlling the belt speed of the one or more conveyor belts by means of the control data. The invention also relates to a container transport device comprising one or more conveyor belts and a control device, wherein the container transport device is configured to carry out a method.

Description

[0001] The invention relates to a method for automatically controlling a container transport device having one or more conveyor belts for adjusting the container density of the containers according to claim 1 and to a container transport device according to claim 15 . Background Art

[0002] DE 195 30 626 A1 discloses a method and a device for detecting the occupancy of a container conveyor having an accumulation zone and / or a deflection zone, wherein the occupancy is detected contactlessly by means of a distance measurement in a direction substantially parallel to the conveying plane, the distance measurement measuring the distance between a reference point upstream of the obstruction and the rearmost container obstructing the container flow.

[0003] In general, the density of containers when transporting them on conveyor belts is not regulated or is insufficiently regulated. This can result in high pressures on the containers, which can lead to glass breakage or abrasions (wear, scratches) and are undesirable when transporting sensitive plastic containers (e.g. containers made of PET) or other thin-walled or sensitive containers such as aluminum cans or tinplate cans. Containers that collide with each other also generate noise and lead to deviations from the desired target distribution of the containers on the conveyor belt.

[0004] Purpose

[0005] The object of the invention is to provide a method and a container transport device which enable gentle and quiet transport of containers on a conveyor belt.

[0006] Solution

[0007] This object is achieved by a method according to claim 1 and by a device according to claim 15. Preferred embodiments are disclosed in the dependent claims.

[0008] A method for automatically controlling a container transport device having one or more conveyor belts in order to adjust the container density of containers transported on the one or more conveyor belts, comprising detecting distance data of containers on the one or more conveyor belts at different times in each case by means of a distance measuring device, transmitting the distance data to a control device by means of the distance measuring device, processing the distance data into control data in the control device, and performing nonlinear control of the belt speed of the one or more conveyor belts by means of the control data.

[0009] The distance measuring device may include various types of sensors which are configured to detect the position of an object, such as a container. This may be achieved optically (photogate) and / or mechanically (blocking switch), by acoustic sensors, cameras, Hall sensors and / or proximity sensors.

[0010] The distance data may include the distance from one container to another container, or the distance from one container to several other containers, and / or the distance from several containers to one container, and / or the distance from several containers to several other containers, and / or the distance from one container to a railing, and / or the distance from several containers to a railing. Alternatively, the distance data may include the number of containers within a defined area and / or the distribution of containers within a defined area.

[0011] An identifier can be assigned to the conveyor belt or to each of the conveyor belts.

[0012] Linear or nonlinear control can be used for control where the output signal is not always proportional to the input signal.

[0013] Here, the nonlinear control is used to control the speed of one or more conveyor belts. If there are multiple conveyor belts, each of these conveyor belts can take into account the speed and direction of travel of the other conveyor belts when controlling its own speed. In addition, the speed can also be controlled in an open loop, so that it can react to changes in the number of containers fed in, so that the container density on multiple conveyor belts can be adjusted at any time.

[0014] The detection of distance data by distance measuring devices may include infrared, sound, sonar and / or laser, lidar and / or mechanical distance measuring devices, such as one or more blocking switches. It is also possible to determine the distance by means of images or video streams, wherein the distance between two or more containers identified in the images is associated with a known container diameter.

[0015] The detection of the distance data may be continuous or metronomic. For example, the distance between a container and at least one adjacent container may be detected, or the distance between several containers may be detected. The distance data may include the information indicated in the above list.

[0016] The nonlinear control may include setting a speed gradient across the plurality of conveyor belts. For example, the speed gradient may be increased when a blockage occurs in the container transport device. For example, the speed gradient may be decreased when the distance exceeds a given distance value.

[0017] A first speed gradient may be set for the conveyor belt moving in a first direction, and a second speed gradient may be set for the conveyor belt moving in a second direction. The first and second speed gradients may be the same or different.

[0018] The gradient can indicate how a variable changes (decreases or increases) over a certain section. In this case, the variable can be considered as the speed of the conveyor belt or the speed of the drive motor, and in this case, the sections can be considered as the first, second, third, etc. conveyor belts. The first conveyor belt can be arranged adjacent to the second conveyor belt, the second conveyor belt can be arranged adjacent to the third conveyor belt, and so on, or the first, second, and third conveyor belts can be connected in series.

[0019] The non-linear control may further comprise adjusting the electrically adjustable guardrails and / or guide plates and / or channel plates of the one or more conveyor belts, and / or adjusting the inclination of the one or more conveyor belts relative to the horizontal plane. For example, in the case where the electrically adjustable guardrails and / or guide plates and / or channel plates are divided into several parts, the adjustment may comprise a segmented adjustment.

[0020] By this adjustment, the space available for containers between the electrically adjustable rails and / or guide plates and / or channel plates of the conveyor belt or conveyors can be increased or decreased, for example. By increasing the adjustment of the available space, additional buffer space can be generated during the operation of the container transport device.

[0021] The non-linear control may further comprise actuating multiple points of one or more resiliently designed deflector plates of the one or more conveyor belts to adjust a profile of the one or more resiliently designed deflector plates.

[0022] By means of such a drive, the containers can be gently transported and / or deflected.

[0023] The method may further comprise detecting current occupancy data of the container occupancy of the one or more conveyor belts by means of one or more position measuring devices. For example, the detection may comprise capturing images using one or more cameras.

[0024] The position device can, for example, detect position data of one or more containers. With the help of the amount of position data, conclusions about the container occupancy of the one or more conveyor belts can be drawn.

[0025] Additionally or alternatively, the position measuring device can detect position data which, for example, also include distance data, as indicated in the above list.

[0026] The method may further include transmitting the current occupancy data to a central computer unit, processing the current occupancy data into processed occupancy data using an algorithm available on the central computer unit, and controlling the drive of the one or more conveyor belts using the processed occupancy data. As an alternative or in addition to the central computer unit, the transmission of the current occupancy data may include transmitting the current occupancy data to a cloud, a smart device, a PLC (programmable logic controller) and / or an IPC (inter-process communication). Alternatively or additionally, the current occupancy data may be detected using a smart device, which may include, for example, a smart camera, and further, the occupancy data may be processed into processed occupancy data, for example, using an algorithm available on the smart device, and the processed occupancy data may be used to control the drive of the one or more conveyor belts.

[0027] For example, the process may be performed continuously.

[0028] For example, these processed occupancy data can be compared with a predetermined container occupancy (target occupancy) of the one or more conveyor belts. If the predetermined occupancy is exceeded, the control can, for example, result in an increase in the drive speed. If the predetermined occupancy is not reached, the control can result in a reduction or increase in the drive speed.

[0029] The method may further comprise processing the current occupancy data using a further algorithm available on the central computer device and calculating therefrom the speed of the one or more conveyor belts using the further algorithm.

[0030] As an alternative or in addition to processing the current occupancy data using the further algorithm available on the central computer unit, the processing of the current occupancy data can be carried out using further algorithms available on the cloud, smart devices, PLCs (programmable logic controllers) and / or IPCs (inter-process communication).

[0031] For example, the method may further comprise calculating an adjustment of an electrically adjustable guardrail and / or guide plate and / or channel plate of the one or more conveyor belts and / or an adjustment of the inclination of the one or more conveyor belts relative to a horizontal plane. For example, in the case where the electrically adjustable guardrail and / or guide plate and / or channel plate are divided into several parts, the adjustment may comprise a segmented adjustment.

[0032] Additionally or alternatively, the method may include predicting and / or learning an adjustment strategy for electrically adjustable guardrails and / or guide plates and / or channel plates of the one or more conveyor belts and / or an adjustment strategy for the inclination of the one or more conveyor belts relative to the horizontal plane. For example, predicting and / or learning may include reaching a predetermined occupancy level for the one conveyor belt or reaching a predetermined occupancy level for the plurality of conveyor belts.

[0033] If the predetermined occupancy is exceeded, it can be arranged that the space between the electrically adjustable rails and / or the guide plates and / or the channel plates of the one or more conveyor belts is increased by the adjustment. If the predetermined occupancy is not reached, it can be arranged that the space between the electrically adjustable rails and / or the guide plates and / or the channel plates of the one or more conveyor belts is reduced by the adjustment.

[0034] The further algorithm may include a neural network or a model-based online optimization. With the model-based online optimization, the model can learn the behavior of the container transporter in a training procedure. With the online optimization, the container transporter can be controlled during operation of the container transporter in the event of deviations from the target behavior (model predictive control).

[0035] The neural network may, for example, include an observation layer, which may be followed by a fully connected layer. A ReLU activation function may then be set four times, each followed by a fully connected layer. The fifth fully connected layer may be followed by a Tanh activation function. The Tanh activation function may be followed by a scaling layer. The output of the neural network may be input to a function to determine the mean square error.

[0036] The output node of the neural network may include a Tanh activation function, through which the suggested action may be mapped to the range [-1, 1].

[0037] This can be useful when the Tanh activation function is followed by a scaling layer. For example, it may be easier to map the output of a neural network to a defined range [min, max] of each belt speed of the one or more conveyor belts.

[0038] For example, the speed of the second conveyor belt may have a limited range of [0.8 0.6] m / s, and the speed of the third conveyor belt may have a limited range of [0.7, 0.5]. Therefore, the output of the Tanh activation function may map values ​​to the range [-1, 1], and in subsequent scaling layers each belt speed action may be adapted to the corresponding limited range.

[0039] The neural network may include at least one hidden layer, and each of the at least one hidden layer may include a ReLU activation function. Among known activation functions, the ReLU activation function is applicable because of its piecewise linear function. For example, the function may directly output the input when the input is positive, and output zero when the input is negative.

[0040] The output of the neural network may include an array. For example, the array may include the corresponding assignment of the identification of the one conveyor belt and its belt speed, or the corresponding assignment of the identification of one conveyor belt among the plurality of conveyor belts and its belt speed.

[0041] The neural network may include at least one fully connected layer, whereby the neural network may learn features from all feature combinations of the previous layer.

[0042] The computation of the neurons of the neural network may depend on the outputs of previous layers, which may have been multiplied by weights and assigned biases.

[0043] To train a neural network, training data can be generated by an agent in a trial and error format. At the beginning of training, the agent can try random actions and the results (reward function) can be evaluated. He can then adjust these actions until he finds the best one. The action can also be added to a noise function in order to allow the agent to explore states that have not yet been explored. The agent can save these states in a buffer of a fixed size. At each iteration, the weights of the neural network can be updated with a mini-batch of data randomly selected from the buffer, so that the data is independently distributed (this is because the agent itself generates the training data).

[0044] The data (which can be an input to a neural network) can be uniform areas that can cover a portion of a conveyor belt so that the occupancy of each of these areas, such as the occupancy rate, can be calculated and it can be determined whether the containers are evenly distributed. If the containers are not evenly distributed, the agent can adjust its actions every n seconds to ensure that the containers are evenly distributed.

[0045] For example, the number of these uniform areas can be 5, 11 or 22, or less or more. The data can also be images or videos from a camera showing the container occupancy of these areas.

[0046] The container transport device comprises one or more conveyor belts and a control device, wherein the container transport device is configured for carrying out the method described above or below. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] For better understanding and description, the accompanying drawings exemplarily show aspects and / or embodiments of the present invention.

[0048] Figure 1 The layers of a neural network are schematically shown;

[0049] Figure 2 schematically showing an oblique view of several conveyor belts and a subsequent bulk transport section having a first number of uniform areas; and

[0050] Figure 3 An oblique view of the several conveyor belts and a subsequent bulk transport section with a second number of uniform areas is schematically shown. DETAILED DESCRIPTION

[0051] Figure 1 A schematic diagram of the layers of a neural network 14 is shown, which may be used for a method and a container transport device as described above or below. Different or similar neural networks may also be used.

[0052] Connected to the observation layer 1 is the first fully connected layer 2, so the neural network 14 can learn features from all feature combinations of the observation layer 1.

[0053] Following the first fully connected layer 2 is the first ReLU activation function 3. For example, the function may directly output the input when it is positive, and output zero when it is negative. The result of the first ReLU activation function 3 is passed to the second fully connected layer 4, so that the neural network 14 can learn features from all feature combinations of the previous layers.

[0054] Connected behind the second fully connected layer 4 is the second ReLU activation function 5. For example, the second ReLU activation function 5 may also directly output the input when the input is positive, and output zero when the input is not positive. The result of the second ReLU activation function 5 is passed to the third fully connected layer 6, so that the neural network 14 can learn features from all feature combinations of the previous layers.

[0055] Connected behind the third fully connected layer 6 is the third ReLU activation function 7. For example, the third ReLU activation function 7 may also directly output the input when the input is positive, and output zero when the input is not positive. The result of the third ReLU activation function 7 is passed to the fourth fully connected layer 8, so that the neural network 14 can learn features from all feature combinations of the previous layers.

[0056] Connected to the fourth fully connected layer 8 is the fifth ReLU activation function 9. For example, the fourth ReLU activation function 9 may also directly output the input when the input is positive, and output zero when the input is not positive. The result of the fourth ReLU activation function 9 is passed to the fifth fully connected layer 10, so that the neural network 14 can learn features from all feature combinations of the previous layers.

[0057] Following the fifth fully connected layer 10 is a Tanh activation function 11. The Tanh activation function 11 can map the proposed action to the range [-1, 1].

[0058] Following the Tanh activation function 11 is a scaling layer 12. The preceding Tanh activation function 11 makes it easier to map the output of the neural network to a limited range [min, max].

[0059] The output of the neural network 14 is input to a function 13 to determine the mean square error.

[0060] Figure 2 Schematically, an oblique view of an apparatus 16 is shown, which comprises a plurality of conveyor belts 17-22, 23-27, 28 and a subsequent bulk transport section 31 having a first number of uniform areas 32-42. The first number is in this case 11. The first group of conveyor belts 17-22 and the single conveyor belt 28 may be drivable in a first direction, and the second group of conveyor belts 23-27 may be drivable in a second direction, wherein the first and second directions are opposite to each other. In the illustration, the bulk transport section 31 may be drivable in a third direction, which is perpendicular to the first and second directions. The bulk transport section 31 may also be drivable in a fourth direction, which is opposite to the third direction.

[0061] Railings 29, 30 are provided above the conveyor belts 17-22, 23-27, 28, which can be used to guide and / or divert containers transported on the conveyor belts 17-22, 23-27, 28. The railings 29, 30 can be electrically adjustable railings. As an alternative or additional solution to the railings, guide plates and / or channel plates can also be provided.

[0062] Figure 3 A schematic oblique view of a device 43 is shown with its conveyor belts 17-22, 23-27, 28 and the subsequent batch transport section 31 as shown in FIG. Figure 2 The conveyor belts 17-22, 23-27, 28 and the batch conveyor 31 are arranged in the same manner. In the figure, the second number of uniform regions 44-65 is 22.

[0063] Since the second number is greater than the first number, for the second device, the container density, ie the occupancy, can be better adjusted using the second number of uniform regions 44-65 in the device.

Claims

1. A method for automatically controlling a container transport device having one or more conveyor belts (17-28) to adjust the container density of containers transported on the one or more conveyor belts (17-28), wherein the method comprises: At different times in each case: - detecting distance data of the containers on the one or more conveyor belts (17-28) by means of a distance measuring device; - transmitting the distance data to a control device via the distance measuring device; - processing the distance data into control data in the control device; - non-linearly controlling the belt speed of the one or more conveyor belts (17-28) by means of the control data.

2. The method according to claim 1, wherein detecting the distance data by the distance measuring device comprises infrared, sound, sonar and / or laser and / or mechanical distance measuring devices, such as one or more blocking switches.

3. The method according to claim 1 or 2, wherein the detection of distance data is performed continuously or metronomically, wherein: For example, the distance between a container and at least one adjacent container is detected, or the distance between multiple containers is detected.

4. The method according to any one of claims 1 to 3, wherein the nonlinear control comprises: A speed gradient is set across the plurality of conveyor belts (17-28), wherein the speed gradient increases, for example, when a blockage occurs in the container transport device, and wherein the speed gradient decreases, for example, when the distance exceeds a given distance value.

5. The method according to any one of claims 1 to 4, wherein the nonlinear control further comprises adjusting electrically adjustable guardrails (29, 30) and / or guide plates and / or channel plates of the one or more conveyor belts, and / or adjusting the inclination of the one or more conveyor belts relative to a horizontal plane, wherein: For example, in the case where the electrically adjustable guardrail (29, 30) and / or the deflector and / or the channel plate are divided into several parts, the adjustment comprises a section-by-section adjustment.

6. The method according to any one of claims 1 to 5, wherein the nonlinear control further comprises driving multiple points of one or more elastically designed deflector plates of the one or more conveyor belts to adjust the profile of the one or more elastically designed deflector plates.

7. The method according to any one of claims 1 to 6, further comprising: - detecting current occupancy data of the container occupancy of the one or more conveyor belts (17-28) by means of one or more position measuring devices, wherein, for example, the detecting comprises recording images with one or more cameras.

8. The method according to claim 7, further comprising: - transmitting said current occupancy data to a central computer unit; - processing the current occupancy data into processed occupancy data using algorithms available on the central computer unit; - using said processed occupancy data to control a drive of said one or more conveyor belts, Therein, for example, the processing is performed continuously.

9. The method according to claim 8, further comprising: - Processing the current occupancy data using a further algorithm available on the central computer unit and calculating therefrom the speed of the one or more conveyor belts (17-28) using the further algorithm, and for example further including calculating the adjustment of electrically adjustable guardrails and / or guide plates and / or channel plates of the one or more conveyor belts, wherein, for example in the case where the electrically adjustable guardrails and / or guide plates and / or channel plates are divided into several parts, the adjustment includes a segment adjustment.

10. The method according to claim 9, wherein the further algorithm comprises a neural network (14) or a model-based online optimization.

11. The method according to claim 10, wherein the output node of the neural network (14) comprises a Tanh activation function (11).

12. The method according to claim 10 or 11, wherein the neural network (14) comprises at least one hidden layer, and each hidden layer of the at least one hidden layer comprises a ReLU activation function, wherein, For example, the ReLU activation function is a piecewise linear function, which directly outputs the input when the input is positive, and outputs zero when the input is not positive.

13. The method according to any one of claims 10 to 12, wherein the output of the neural network (14) comprises an array, wherein: For example, the array includes an identification of a conveyor belt and a corresponding assignment of its belt speed, or an identification of a conveyor belt among the plurality of conveyor belts and a corresponding assignment of its belt speed.

14. The method according to any one of claims 10 to 13, wherein the neural network (14) comprises at least one fully connected layer (2, 4, 6, 8, 10), whereby the neural network (14) learns features from all feature combinations of the previous layer.

15. A container transport device comprising one or more conveyor belts (17-28) and a control device, wherein the container transport device is configured for carrying out the method according to any one of claims 1 to 14.

Images