Method for monitoring the vulcanization process of vehicle tires and evaluation device

Abstract

The invention relates to a method for monitoring a vulcanization process of a vehicle tire and an evaluation device, wherein at least one sensor value and / or at least one control variable of a tire press apparatus (100) is detected, wherein the at least one sensor value and / or the at least one control variable is also fed to the evaluation device, the evaluation device comprising an ML model set by means of a machine learning method, and wherein the evaluation device and / or the ML model outputs a warning message if an evaluation of the at least one sensor value and / or the at least one control variable by the evaluation device leads to the conclusion that a tire press apparatus (100) or at least one part of a tire press apparatus (100) is faulty or abnormal, or that a fault or abnormality of a tire press apparatus (100) or at least one part of a tire press apparatus (100) will occur in a foreseeable time.

Description

TECHNICAL FIELD

[0001] The present application relates to a method and an evaluation device for monitoring a vulcanization process of a vehicle tire in a tire heat pressing apparatus, wherein the tire heat pressing apparatus comprises in particular an elastic heat bladder, which is designed and configured to support the vulcanization process of a vehicle tire located in the tire heat pressing apparatus, and wherein at least one sensor value and / or at least one control variable of the tire heat pressing apparatus is detected. BACKGROUND

[0002] Such a method is known from the prior art. Thus, for example, the patent document US 7, 987, 697 B2 discloses a method for process control of a vulcanization process of a vehicle tire, wherein nitrogen is additionally mixed to the gas used to inflate the heat bladder in the context of the vulcanization process. If the heat bladder now has a leak, nitrogen can also be proven in the area between the heat bladder and the tire to be vulcanized. With the aid of a corresponding proving system, a leak in the heat bladder can be proven in this way.

[0003] The disadvantage of the prior art is that the tire heat pressing apparatus has to be expanded in terms of structure and equipment in order to be able to prove a corresponding leak in the heat bladder of the press. In addition, further gases are additionally required in addition to the normal process gases, which further increases the effort and costs of the corresponding tire heat pressing apparatus. SUMMARY

[0004] It is therefore an object of the present application to provide a method or a system which reduces the effort of equipment and / or structure for implementing a facility monitoring and / or process control of a tire heat pressing apparatus relative to the prior art.

[0005] The method is designed and configured for monitoring a vulcanization process of a vehicle tire in a tire heat pressing apparatus,

[0006] wherein at least one sensor value and / or at least one control variable of the tire heat pressing apparatus is detected, and wherein, in addition,

[0007] at least one sensor value and / or at least one control variable is delivered to an evaluation device for evaluation,

[0008] the evaluation device comprises an ML model configured by means of a machine learning method,

[0009] and if an evaluation of the at least one sensor value and / or the at least one control variable by the evaluation device leads to

[0010] o a fault or an anomaly of the tire heat pressing apparatus or at least one part of the tire heat pressing apparatus,

[0011] or the tire heat pressing apparatus or at least one part of the tire heat pressing apparatus will fail or will have an anomaly in a foreseeable time,

[0012] then the evaluation device and / or the ML model outputs a warning message.

[0013] Here it can be proposed that if the evaluation device derives from the ML model

[0014] - the tire heat pressing apparatus or at least one part of the tire heat pressing apparatus has a failure or an anomaly,

[0015] - or the tire heat pressing apparatus or at least one part of the tire heat pressing apparatus will fail or will have an anomaly in a foreseeable time,

[0016] only then a warning message is output.

[0017] The method can be designed and set up, for example, such that only sensor values of sensors that are already present in the tire heat pressing apparatus overall are used. Control variables can also be obtained, for example, from a control device of the tire heat pressing apparatus, which is standardly provided in the tire heat pressing apparatus. In this way, the equipment and / or structural expenditure for implementing a process control in the tire heat pressing apparatus is significantly reduced with respect to the prior art, since, for example, at least no separate hardware is required for this. The method according to the present description can be implemented more easily, for example, in an existing tire heat pressing apparatus.

[0018] The use of sensors that are already present in the tire heat pressing apparatus, for example, is further simplified by using the ML model in that such an ML model is trained accordingly or set up accordingly such that the recognition of, for example, when a respective disturbance or anomaly occurs or can occur can be matched to the special characteristics of the constructed sensor system. This also contributes to further reducing the equipment and / or structural expenditure for implementing a process control for the tire heat pressing apparatus.

[0019] In an advantageous design variant, it can be proposed, for example, that at least one sensor value and / or at least one control variable is detected in the context of a vulcanization process of a vehicle tire located in the tire heat pressing apparatus. Here, the vulcanization process can comprise the following steps, for example: loading the tire heat pressing apparatus with a tire blank of a vehicle tire, vulcanizing the vehicle tire and removing the vulcanized vehicle tire from the tire heat pressing apparatus.

[0020] It can further be proposed that the at least one sensor value and / or the at least one control variable is detected in the context of a downtime, a preparation time and / or a post-processing time of the tire heat pressing apparatus. Here, the downtime of the tire heat pressing apparatus can be designed and set, for example, as the time in which no vehicle tire is vulcanized in the tire heat pressing apparatus. The preparation time can be designed and set, for example, such that the tire heat pressing apparatus for vulcanizing a vehicle tire is prepared in this time. The post-processing time can be designed and set such that the tire heat pressing apparatus transitions from vulcanizing a vehicle tire to the downtime in the post-processing time.

[0021] It can further be proposed that the at least one sensor value and / or the at least one control variable is detected in the context of a separate analysis method sequence of the heat pressing apparatus. Here, the analysis method sequence can be designed and set such that an anomaly or a fault of the tire heat pressing apparatus or a part thereof can be particularly advantageously ascertained (or predicted) by evaluating the at least one sensor value or the at least one control variable by means of the evaluation device, for example, using an ML model.

[0022] The tire heat pressing apparatus, also referred to as vulcanization press or "vulcanizer", is understood to be a machine or a facility which is designed and set up for vulcanizing vehicle tires. This method step is a hardening step in the context of producing vehicle tires, in which the respective vehicle tire obtains its final shape. In the hardening process, the vehicle tire is vulcanized at an appropriate pressure and at an appropriate temperature for a certain time. During this working step, the raw rubber is converted into a flexible and elastic rubber. Furthermore, the tire obtains its contour and its sidewall markings in the respective mold of the vulcanization press or the tire heat pressing apparatus.

[0023] Such a vulcanization press usually comprises one or two molds, which can accommodate each one so-called tire blank or "green tire". Such a vulcanization press usually comprises a line system for hot gases, liquids or vacuum and a transport belt for transporting and removing the tire blanks and the vulcanized vehicle tires.

[0024] The so-called heat bladder is then usually positioned in the inner region of the tire blank in the context of the vulcanization process of the tire blank in such a tire heat pressing apparatus, which is similar to a bicycle inner tube, and hot gases or hot liquids are then delivered to the heat bladder at high pressure. In this way, the heat bladder presses the tire blank into the mold, in which the tire contour and markings are then, for example, imprinted, and the vulcanization process of the tire is also triggered by the high temperature. After the vulcanization process, the vulcanized vehicle tire is now removed from the respective mold and delivered to the next process step of tire production.

[0025] For example, the at least one sensor value can be a digital or analog value provided by the sensor. Here, the value can be provided, for example, directly by the sensor or can also be provided after a corresponding pre-processing or digital-analog conversion or analog-digital conversion. Here, the sensor can be any type of sensor, i.e. for example a temperature sensor, a pressure sensor, a flow sensor, a fill level sensor, a speed sensor, a counter, an acceleration sensor or a similar sensor. Furthermore, the sensor can also be designed and configured as a so-called software sensor or virtual sensor.

[0026] The at least one sensor value can be, for example, a value detected by the sensor at a specific point in time, for example in the context of a vulcanization process or also in the context of a specific analytical method procedure. Furthermore, the at least one sensor value can be designed and configured as a plurality of sensor values, wherein each of the plurality of sensor values can originate from a different sensor and be detected, for example, at a specific, for example predetermined or assignable, point in time, for example in the context of a vulcanization process of a specific analytical method procedure.

[0027] Furthermore, the at least one sensor value can be designed and configured as one or more time series values of one or more sensors, wherein, for example, a time series of a sensor can be recorded at regular points in time or also at specific, fixedly preset points in time with different time intervals, for example in the context of a vulcanization process or a specific analytical method procedure.

[0028] The at least one sensor value can also comprise a combination of values according to the present specification, for example a combination of values of one or more sensors according to the present specification.

[0029] The control variable of the tire heat pressing apparatus can be, for example, a digital or analog value, a command, a command sequence or also a message, which is transmitted, for example, from a control device to an actuator of the tire heat pressing apparatus or to another control device. Here, the control device can be, for example, part of the tire heat pressing apparatus or also outside the tire heat pressing apparatus. Here, the control device can be, for example, part of the tire heat pressing apparatus or also outside the tire heat pressing apparatus. The actuator can be, for example, a motor, a drive, a regulator, a valve, a display device, a pump, a valve regulator or a similar device.

[0030] The control device can be, for example, a computer, a programmable logic controller, a modular programmable logic controller, a controller, a microcontroller, a decentralized peripheral device or a similar device.

[0031] Here, the analog or digital values or commands can be transmitted from the control device to the actuator of the tire heat pressing apparatus or to another control device, for example, via a single-wire line, a two-wire line, a field bus or a similar wired or wireless communication connection.

[0032] Very generally, the control device can be any type of computer or computer system, which is designed and set up, for example, for controlling an instrument, a device, a plant or a facility. The controller can also be a computer, a computer system or a so-called "cloud", on which a control software or a control software application, for example a control application or a control App, is implemented or installed. Such a control application implemented in the cloud, for example, can be designed and set up as an application with the functionality of a programmable logic controller.

[0033] The control device can also be designed and set up as a so-called edge device, wherein such an edge device can comprise an application for controlling a device or a facility, for example. Such an application can be designed and set up as an application with the functionality of a programmable logic controller, for example. Here, the edge device can be connected to another control device of a device or a facility, for example, or can also be connected directly to the device or the facility to be controlled. Furthermore, the edge device can be designed and set up such that the edge device is additionally connected to a data network or a cloud or is designed and set up for connection to a respective data network or a respective cloud.

[0034] The control device can also be designed and set up as a so-called programmable logic controller (SPS), for example. Furthermore, the control device can also be set up and designed as a so-called modular programmable logic controller (modular SPS).

[0035] A programmable logic controller, in short SPS, is a programmed and used for regulating or controlling a component of a facility or a machine. Specific functions, i.e. process control, for example, can be implemented in the SPS, so that inputs and output signals of a process or a machine can be controlled in this way. The programmable logic controller is defined at least partially in the EN 61131 standard or the IEC 61131 standard, for example.

[0036] In order to couple the programmable logic controller to the facility or the machine, actuators and sensors are used, which are usually connected to the outputs of the programmable logic controller. Furthermore, status displays are used. Basically, sensors are located at the inputs of the SPS, wherein the programmable logic controller obtains information about what is happening in the facility or the machine via the inputs. Gratings, limit switches, pushbuttons, incremental encoders, level sensors, temperature probes and pressure sensors are suitable as sensors, for example. Contactors for switching on electrical motors, electric valves for compressed air or hydraulic devices, drive control modules, motors, drives are suitable as actuators, for example.

[0037] The SPS can be realized in different ways and methods. That is to say, the SPS can be realized as an electronic stand-alone device, a software simulation, a PC plug-in card, etc. Usually, there are also modular solutions, in the context of which the SPS consists of a plurality of plug-in modules.

[0038] The evaluation device can be any computer device having sufficient memory and computing power to undertake the storage, execution and / or operation of a corresponding ML model. In particular, the evaluation device can be, for example, a corresponding control device, a corresponding controller, a programmable logic controller, a modular programmable logic controller or a similar device. Furthermore, the evaluation device can be designed and set up, for example, as a so-called edge device. The evaluation device can also be part of a computer hardware, a computer or a computer network, for example an application in the cloud or an application in a computer network or an application in a corresponding server device or computer device.

[0039] The evaluation device can also be designed and set up, for example, as a virtual evaluation device entity, which is running or can be running, for example, on a corresponding hardware, a corresponding computer network or a cloud. Here, the functions of the evaluation device are generated in the running of the virtual evaluation device instance.

[0040] For example, the evaluation device can also be part of a control device or a control computer for a tire heat pressing apparatus.

[0041] The evaluation device can also consist of a plurality of sub-devices. For example, the sub-devices can be communicatively coupled. Here, each sub-device can then be designed and set up according to the design options explained above for the evaluation device.

[0042] The evaluation device can also be designed and set up such that at least one sensor value delivered thereto and / or at least one control variable delivered thereto is transmitted directly or indirectly to the ML model, in particular as an input variable to the ML model.

[0043] Furthermore, the evaluation device can be designed and set up for processing or further processing at least one sensor value and / or at least one control variable. Such processing or further processing can be, for example, or include normalization, scaling, transformation, reformatting, translation and / or other similar processing steps.

[0044] Here, the evaluation device can be designed and set up such that after processing or further processing at least one sensor value and / or at least one control variable, the result(s) of such processing or further processing are delivered to the ML model, in particular as input variable(s) to the ML model.

[0045] The evaluation device can also include, for example, a simulation environment designed and set up for running a simulation program for simulating a tire heat pressing apparatus or for simulating parts or components of a tire heat pressing apparatus. The evaluation device can also include a simulation environment with a simulation program for simulating a tire heat pressing apparatus or also for simulating one or more parts or components of a tire heat pressing apparatus.

[0046] The simulated part or component of the tire heat pressing apparatus can for example be or comprise an apparatus component which is provided for conveying and / or conducting a gas or a liquid to the elastic heat bladder. Furthermore, the simulated part of the tire heat pressing apparatus can also be or comprise the elastic heat bladder itself.

[0047] Here, the processing or further processing of the at least one sensor value and / or of the at least one control variable by the evaluation device in the proposed manner can be designed and configured to use the at least one sensor value or the at least one control variable as input information or input data for a simulation program according to the present specification which is run in a simulation environment according to the present specification.

[0048] The result or results of the thus designed processing or further processing of the at least one sensor value and / or of the at least one control variable can then for example be one or more values which are generated in the context of the simulation program run, which values can then for example be further conveyed to the ML model, in particular as input variables to the ML model. The one or more such values generated in the context of the simulation program run can for example be designed and configured to be output values of so-called virtual sensors (also referred to as so-called "soft sensors") simulated within the simulation. Examples of such soft sensors can for example be virtual pressure sensors, virtual vapor sensors, virtual temperature sensors and / or virtual sensors for measuring material or energy flux.

[0049] Here, for example, the proposed simulation can be designed and configured such that it runs in parallel to a vulcanization process or a part thereof running in the tire heat pressing apparatus.

[0050] Furthermore, it can be proposed that the at least one sensor value conveyed to the evaluation device and / or the at least one control variable conveyed to the evaluation device are used as input variables for the simulation, for example a simulation according to the present specification.

[0051] Furthermore, it can be proposed that one or more values generated by the simulation, in particular a simulation according to the present specification, are used as input variables for the ML model. For example, it can thus be proposed that one or more so-called "soft sensors" are defined within such a simulation and the respective soft sensor values are then calculated and shown in the context of the simulation run. Here, it can also be proposed that such soft sensor values of one or more such soft sensors are used as input variables for the ML model. Examples of soft sensors can for example be virtual pressure sensors, virtual vapor sensors, virtual temperature sensors and / or virtual sensors for measuring material or energy flux.

[0052] In this way, a further advantageous design is also achieved, because by using such virtual sensors it is possible to generate and use or use multiple sensors for the ML model, without having to expand the tire heat press device in terms of structure and equipment.

[0053] Machine learning methods are understood, for example, as an automated (“machine”) method which does not generate a result by means of predetermined rules, but in which regularities are recognized from a plurality of instances, usually automatically, by means of a machine learning algorithm or learning method, and a report about the data to be analyzed is then generated on the basis of the regularities.

[0054] Such machine learning methods can be designed and set up, for example, as a supervised learning method, a partially supervised learning method, an unsupervised learning method or a reinforcement learning method (“Reinforcement Learning”).

[0055] Examples of machine learning methods include, for example, a regression algorithm (for example a linear regression algorithm), the generation or optimization of decision trees (so-called “Decision Trees”), a learning method for neural networks, a clustering method (for example the so-called “k-means clustering”), a learning method for or to generate support vector machines (“Support Vector Machines” (SVM)), a learning method for or to generate sequential decision models or a learning method for or to generate Bayesian models or networks.

[0056] The result of such an application of a machine learning algorithm or learning method to specific data is referred to in particular as a “machine learning” model or ML model in this specification. Here, such a ML model again represents the digitally stored or storable result of the application of a machine learning algorithm or learning method to analyzed data.

[0057] Here, the generation of a ML model can be designed and set up in such a way that the ML model is reformed by using a machine learning method or that an already existing ML model is changed or adapted by using a machine learning method.

[0058] Examples of such ML models are the result of a regression algorithm (for example a linear regression algorithm), of a neural network (“Neural Networks”), of a decision tree (“Decision Tree”), of a clustering method (including, for example, obtained clusters or categories of clusters, limits and / or parameters), of a support vector machine (“Support Vector Machines” (SVM)), of a sequential decision model or of a Bayesian model or network.

[0059] Here, the neural network can be, for example, a so-called "deep neural network", a "feedforward neural network", a "recurrent neural network", a "convolutional neural network" or an "autoencoder neural network". Here, the application of the respective machine learning method to the neural network is also generally referred to as "training" the respective neural network.

[0060] The decision tree can be designed and set up, for example, as a so-called "iterative dichotomiser 3" (ID3), classification or regression tree (CART) or a so-called "random forest".

[0061] At least in conjunction with the present description, a neural network is understood to be an electronic device comprising a network of so-called nodes, wherein each node is generally connected to a plurality of other nodes. A node is also referred to as neuron, cell or unit, for example. Here, each node has at least one input connection and one output connection. An input node for a neural network is understood to be a node which can receive signals (data, stimuli, patterns, etc.) from the outside world. An output node of a neural network is understood to be a node which can forward signals, data, etc. to the outside world. A so-called "hidden node" is understood to be a node of a neural network which is neither configured as an input node nor as an output node.

[0062] Here, the neural network can be configured as a so-called deep neural network (DNN), for example. Such a "deep neural network" is a neural network in which the network nodes are arranged in layers (where the layers themselves can be one-, two- or even higher-dimensional). Here, a deep neural network comprises at least one or two so-called hidden layers, which layers only comprise nodes which are not input nodes or output nodes. That is to say, the hidden layers do not have a connection to an input signal or an output signal.

[0063] Here, so-called "deep learning" is understood to be a class of machine learning techniques which use multilayer nonlinear information processing for supervised or unsupervised feature extraction and feature transformation and for pattern analysis and pattern classification, for example. Here, within the scope of such "deep learning", so-called "deep neural networks" are generally used, for example according to the present description.

[0064] The neural network can also have a so-called autoencoder structure, for example, which is explained in more detail in the description of the present description. Such an autoencoder structure can be suitable for reducing the dimensionality of data, for example, and for identifying similarities and commonalities, for example.

[0065] The neural network can also be configured as a so-called classification network, for example, which is particularly suitable for dividing data into classes. Such a classification network is used in conjunction with handwriting recognition, for example.

[0066] Another possible structure of a neural network can for example be a design scheme as a so-called "deep belief network".

[0067] For example, a neural network can also have a combination of several of the above-mentioned structures. Thus, for example, the architecture of a neural network can comprise an autoencoder structure to reduce the dimensionality of input data, which input data can then be combined with another network structure in order to, for example, identify particularities and / or anomalies within the reduced dimensionality of data or to classify the reduced dimensionality of data.

[0068] The values describing the individual nodes and their connections can comprise further values describing a particular neural network, for example, can be stored in a set of values describing the neural network. Such a set of values then represents, for example, a design scheme of the neural network. If such a set of values is stored after the training of the neural network, the design scheme of the trained neural network is thereby stored, for example. Thus, for example, a neural network can be trained in a first computer system with corresponding training data, the corresponding set of values associated with the neural network is then stored and transmitted as a design scheme of the trained neural network to a second system, and the trained neural network is then used there.

[0069] A neural network can generally be trained by inputting input data into the neural network and then analyzing the corresponding output data from the neural network via various known learning methods to determine parameter values for the individual nodes or for their connections. In this way, a neural network can be trained with known data, patterns, stimuli or signals in a manner known today in order to be able to use the thus trained network for analyzing other data, for example, later on.

[0070] Generally, training a neural network is understood to mean processing data for training the neural network in the neural network by means of one or more training algorithms in order to calculate or change the so-called bias values ("bias biases"), weight values ("weights weights") and / or transfer functions ("Transfer Functions") of the individual nodes of the neural network or of the connections between two nodes within the neural network.

[0071] For training the neural network, for example, according to the present specification, for example one of the so-called "supervised learning" methods can be used. Here, by means of the training with the respective training data, the network is trained with the results or capabilities respectively associated with the data. In addition, it is also possible to train the neural network using the so-called "unsupervised learning" method. For example, such an algorithm generates a model describing the input and from which a prediction is implemented for a given quantity of data. Here, for example, there are clustering methods, if the data are distinguished from one another by characteristic patterns, the data can be divided into different categories by means of the clustering methods.

[0072] In training the neural network, the supervised and unsupervised learning methods can also be combined, for example, if a part of the data is associated with a trainable characteristic or capability, while this is not the case in another part of the data.

[0073] In addition, the so-called "reinforcement learning" method can also be used at least for training the neural network.

[0074] For example, the training, which requires a relatively high computing power of the respective computer, can be carried out on a high-performance system, while with the trained neural network other work or data analysis can then be carried out completely on a low-performance system. For example, such other work and / or data analysis can be carried out with the trained neural network on a secondary system and / or on a control device, programmable logic controller or modular programmable logic controller or other respective device according to the present specification.

[0075] In the context of the present specification, a fault is understood to be in particular each faulty function of a device or apparatus, here a tire heat press. In particular, a fault is considered to be such a faulty function which leads to a faulty product design, an abnormal wear or also a damage to the device or apparatus or its components.

[0076] In the context of the present specification, an anomaly is understood to be, for example, that the respective device or the respective apparatus does not behave as intended. Here, the presence of an anomaly does not necessarily lead to an error in the manufactured product, here for example a product manufactured by means of a tire heat press, but for example can be a behavior of the apparatus or device which was not foreseen when setting up or planning the device or apparatus. Since such an anomaly can lead to unforeseeable behavior of the device or apparatus, which in turn can lead to damage to the device, destruction or endangerment of persons and materials, the recognition of such an anomaly is also helpful or often even necessary for the safe operation of the device or apparatus.

[0077] For example, the fault or the anomaly will occur in a foreseeable time, for example, it is possible to design and set up such that the time period detected with the "in a foreseeable time" is a time period which is at least essentially necessary for the organization to eliminate the cause of the fault or the anomaly and to find and determine the cause. Furthermore, the time period detected with the "in a foreseeable time" can also be a time period which is necessary or can be necessary for planning a corresponding processing work for eliminating the cause and / or the fault or the anomaly. Furthermore, the time period detected with the "in a foreseeable time" can also be a time period until the next planned maintenance period or until the next planned maintenance time period. Furthermore, the time period detected with the "in a foreseeable time" can also comprise a plurality of the above-mentioned time periods.

[0078] The warning message can comprise, for example, information about the identified anomaly and the identified possible fault and / or disturbance, or also information about a possible anomaly, fault and / or disturbance which will occur. Furthermore, the warning message can also comprise information about a possible cause of the identified fault and / or the identified anomaly and options for finding such a cause. Furthermore, such a warning message can also comprise information about a person, a corresponding group of persons or also a corresponding service device which is responsible for a corresponding maintenance or elimination of the cause of the possible fault, and corresponding contact data, i.e. for example, an address, a telephone number, an e-mail address, etc.

[0079] If the evaluation of the at least one sensor value and / or the at least one control variable results in the existence of a fault or an anomaly or a fault or an anomaly will occur in a foreseeable time, the evaluation device or the evaluation device with the ML model outputs a warning message, which can be designed and set up in the manner explained below.

[0080] In the case of the ML model being configured as a neural network, for example, it can be provided that, after the at least one sensor value and / or the at least one control variable is input into the neural network, the neural network outputs, for example, the corresponding information. The neural network, if necessary after a further evaluation step by the evaluation device, displays the output of the corresponding warning message.

[0081] Here, the neural network can be trained, for example, using a monitoring learning method by means of corresponding sensor values and / or control variables of the tire heat pressing apparatus and / or the hot bladder of the tire heat pressing apparatus and known states. This is explained in more detail in the description below. In this case, the evaluation device can be designed and set up such that the information is output by a corresponding output node layer of the neural network, by means of which the output of the corresponding warning message is displayed.

[0082] Moreover, it is also possible to train the neural network using a monitor-less learning method with the aid of corresponding sensor values and / or control variables. This is also explained in more detail below. If, in this case, for example, a neural network having a so-called autoencoder structure is used, as is explained in more detail in the further description, it is thus possible, for example, to propose that a change in the so-called code node layer of the autoencoder (as explained in more detail below) displays the output of a corresponding warning message.

[0083] In the case of the design variants for the ML model as explained in more detail below, the procedure for displaying the warning message output can be implemented in an analogous manner using the ML model.

[0084] In an advantageous design variant, the tire heat pressing apparatus can comprise an elastic heat bladder, the heat bladder being designed and configured to support a vulcanization process of a vehicle tire located in the tire heat pressing apparatus.

[0085] Here, the tire heat pressing apparatus with the heat bladder can be designed and configured to support a vulcanization process of a vehicle tire located in the tire heat pressing apparatus using a gas that is fed or can be fed to the heat bladder or a liquid that is fed or can be fed to the heat bladder.

[0086] Moreover, it can be proposed that, in order to support a vulcanization process of a vehicle tire located in the tire heat pressing apparatus, a hot or heated gas or a hot or heated liquid is fed or can be fed to the heat bladder.

[0087] The heat bladder (also referred to in English as "Bladder" or "Tire Curing Bladder") is designed and configured as an elastic bubble (for example made of rubber or a similar material) that is inflated or expanded by means of a gas or a liquid being fed and that can be depressurized by means of the gas contained therein or the liquid contained therein being blown out and / or sucked out. Here, the heat bladder can be designed and configured, for example, such that it can expand or expand when filled with a gas or a liquid and correspondingly contract or contract again when emptied.

[0088] Here, the heat bladder can also be designed and configured, for example, such that the feedable or fed gas or the fed or feedable liquid can have a temperature of up to 150 degrees Celsius, advantageously up to 200 degrees Celsius, and also advantageously up to 250 degrees Celsius.

[0089] The heat bladder can also be designed and configured such that, for example, air, steam, nitrogen and / or other gases (also having the above-mentioned temperatures) can be fed as a gas. Water is typically used as a liquid, for example.

[0090] Here, the heat capsule can be designed and configured in such a way that it can withstand a pressure of up to 30 bar or more, or also a negative pressure of up to -1 bar, at least when it is located in the tire blank.

[0091] The monitoring of the heat capsule plays a particular role in the context of monitoring the vulcanization process in the tire hot press apparatus. Thus, for example, cracks and / or leaks in the heat capsule material can lead to hot gases or hot liquids entering the intermediate space between the heat capsule and the profile blank, and to the vehicle tires produced then becoming faulty and / or of reduced quality. If such a crack or failure in the heat capsule is identified too late, for example, this can lead to a greater number of vehicle tires of reduced quality or even a greater number of unusable vehicle tires being produced in the tire hot press apparatus.

[0092] It is therefore advantageous to identify such damage in the heat capsule early on, or even better to have determined that such damage will occur in the near future before a real failure has occurred. This can be done, for example, by identifying a corresponding precursor of such a failure, for example a corresponding micro-crack.

[0093] Furthermore, the evaluation device and / or the ML model can be designed and configured in such a way that, if the evaluation of the at least one sensor value or of the at least one control variable by the evaluation device leads to:

[0094] - the heat capsule is faulty or abnormal,

[0095] - or the heat capsule will fail or be abnormal in a foreseeable time,

[0096] a warning warning message is output. Here, the evaluation of the at least one sensor value or of the at least one control variable by the evaluation device can advantageously be carried out using an ML model.

[0097] In this advantageous design, it can be proposed, for example, that the method is designed and configured in such a way that, if the evaluation of the at least one sensor value or of the at least one control variable by the evaluation device using the ML model leads to the heat capsule being faulty or to a failure of the heat capsule in a foreseeable time, a warning message is output.

[0098] The above object is therefore also achieved by a method for monitoring a vulcanization process of a vehicle tire in a tire hot press apparatus having an elastic heat capsule, wherein the elastic heat capsule is designed and configured to support the vulcanization process of a vehicle tire located in the tire hot press apparatus. In the context of the method, at least one sensor value and / or at least one control variable of the tire hot press apparatus is detected, wherein the at least one sensor value and / or the at least one control variable is fed to an evaluation device for evaluation,

[0099] wherein the evaluation device comprises an ML model set by means of a machine learning method, and wherein the evaluation device and / or the ML model outputs a warning message if an evaluation of the at least one sensor value and / or of the at least one control variable by means of the ML model by the evaluation device leads to:

[0100] - a malfunction or an anomaly of the heat bladder,

[0101] - or a malfunction or an anomaly of the heat bladder will occur in a foreseeable time,

[0102]

[0103] Since the heat bladder is a wearing part of the tire heat pressing apparatus and thus relatively often involves corresponding anomalies and malfunctions, the process control or the facility monitoring for the tire heat pressing apparatus can be set particularly effectively by means of the design. Furthermore, a functioning in accordance with regulations of the heat bladder is also important for the quality of the vehicle tires produced and for a corresponding resource-saving operation of the tire heat pressing apparatus.

[0104] The malfunction or the anomaly of the heat bladder or the impending malfunction or the impending anomaly of the heat bladder can be manifested in various parameters of the tire heat pressing apparatus or in parameter curves over time, wherein each corresponding parameter can also exhibit anomalies for other reasons, if necessary.

[0105] Thus, it is generally possible to deduce an impending anomaly or an impending malfunction of the heat bladder from specific parameters or specific changes of parameters of the tire heat pressing apparatus in the first place. Since this correlation is often relatively complex or relatively difficult to analyze, the effort for identifying this malfunction and anomaly or the impending effect and anomaly can be reduced here advantageously by using the ML model.

[0106] The method can be designed and set here such that all of the at least one sensor value is provided by such a sensor of the tire heat pressing apparatus, which is necessary for the preset vulcanization process of the vehicle tire in the vehicle tire heat pressing apparatus or is used for the preset vulcanization process of the vehicle tire in the vehicle tire heat pressing apparatus.

[0107] The advantage of this design is that no special precautions have to be taken in the context of the sensor device for the operation of the evaluation device or for identifying whether a malfunction or an anomaly is present or whether a malfunction or an anomaly will occur in a foreseeable time, but only the sensors that are generally preset for the corresponding production of vehicle tires with a preset quality are used.

[0108] This approach can further reduce the effort in terms of equipment and / or structure for implementing the facility monitoring or the process control for the tire heat pressing apparatus.

[0109] ​For example, if a sensor is required in the context of a vulcanization process in order to produce a tire of a preset quality, a sensor is required for the preset vulcanization process. Such a "required" sensor is not only set up for example for predictive maintenance or for determining a facility failure or wear of a facility or a component thereof. Furthermore, a "required" sensor can also be a sensor which is not only set up for use with an evaluation device according to the present description.

[0110] For example, when a sensor is used in the context of a flow of a vulcanization process of a vehicle tire in order to produce a tire of a preset or presettable quality, the sensor is used for the preset vulcanization process. Such a "used" sensor is not only set up for example for predictive maintenance or for determining a facility failure or wear of a facility or a component thereof. Furthermore, a "used" sensor can also be a sensor which is not only set up for use with an evaluation device according to the present description.

[0111] For example, a sensor which is required or used for a preset vulcanization process of a vehicle tire in a tire press means that if an evaluation device according to the present application (or a device similar thereto) is not set up in the tire press, a corresponding sensor is also present.

[0112] For example, a sensor which is required or used for a preset vulcanization process of a vehicle tire in a tire press also means that if the tire press is designed and set up only for vulcanizing vehicle tires and is not used for carrying out or supporting a method for facility monitoring, for predictive maintenance, for quality control and / or does not run or support an evaluation device according to the present description, a corresponding sensor is also present.

[0113] Furthermore, the method according to the present description can be designed and set up such that the tire press comprises a control device for controlling the vulcanization process and such that at least one of the at least one control variable comprises a control variable output by the control device for the tire press or a component of the tire press in the context of the vulcanization process.

[0114] By using a corresponding control variable for a tire press in the context of the method according to the present description, it is possible to reduce the operational or structural effort for implementing a facility monitoring or process control, since such a control variable is already required individually for the operation of the corresponding tire press.

[0115] Here, the actuating variable for the tire heat pressing apparatus can be any electrical or electronic control variable that is transmitted from the controller or control device to the tire heat pressing apparatus or components thereof. In particular, the actuating variable can be any of the following electrical or electronic control variables that are transmitted from the controller or control device to the tire heat pressing apparatus or components thereof in the context of the vulcanization process of the vehicle tire within the tire heat pressing apparatus. Here, the actuating variable can be transmitted, inter alia, for example, to an actuator of the tire heat pressing apparatus.

[0116] The electrical control variable can be, for example, an analog signal, for example, a corresponding pulse, a voltage level, a signal of a specific frequency or also a modulated signal of a specific frequency or similar. The electronic control variable can be any type of digital information, for example, one or more digital variables or digital messages.

[0117] The method or tire heat pressing apparatus according to the present description can also be designed and configured such that the tire heat pressing apparatus comprises a pressure sensor and at least one of the at least one sensor value is provided by the pressure sensor,

[0118] and / or

[0119] The tire heat pressing apparatus comprises a temperature sensor and at least one of the at least one sensor value is provided by the temperature sensor.

[0120] Here, the temperature sensor can be any type of sensor that measures the temperature within the tire heat pressing apparatus or one or more components of the tire heat pressing apparatus and is transmitted, for example, to a control device of the tire heat pressing apparatus and / or to an evaluation device. Here, for example, the temperature of a component of the tire heat pressing apparatus can be measured or also the temperature of a process gas or process liquid or other process substances. Furthermore, for example, the temperature of a vehicle tire or of a vehicle tire to be vulcanized or vulcanized can also be measured during the vulcanization process.

[0121] Here, the pressure sensor can be any type of sensor that measures the pressure within the tire heat pressing apparatus or one or more components of the tire heat pressing apparatus and is transmitted, for example, to a control device of the tire heat pressing apparatus and / or to an evaluation device. Here, for example, the pressure of a process gas or process liquid or other process substances can be measured. For example, the pressure within a delivery line or discharge line of the tire heat pressing apparatus, for example, for a process substance used in the context of the vulcanization process, can be measured by the pressure sensor. In a preferred design, the pressure in a heat capsule of the tire heat pressing apparatus can also be measured, for example, by the pressure sensor.

[0122] Furthermore, the method according to the present specification and / or the tire heat pressing apparatus according to the present specification can be designed and configured such that the tire heat pressing apparatus comprises at least one delivery valve for regulating the delivery of the gas or liquid to the heat capsule and at least one discharge valve for regulating the removal of the gas or liquid from the heat capsule, and

[0123] At least one of the at least one sensor values is provided by a position sensor of the at least one delivery valve and / or by a position sensor of the at least one discharge valve,

[0124] and / or

[0125] At least one of the at least one control variable comprises a drive control variable for the at least one delivery valve and / or a drive control variable for the at least one discharge valve.

[0126] Here, the delivery valve and / or the discharge valve can have, for example, two states, in particular exactly two states: a closed state, in which the process substance is prevented from flowing through the valve, and an open state, in which the process substance flows through the valve. Furthermore, the delivery and / or discharge valve can also be designed and configured such that it can continuously regulate the flow rate of the respective process substance.

[0127] Each of the mentioned valves can have, for example, a position sensor, from the signal of which the open state or the flow state of the valve can be determined, for example. Such a position sensor can be realized, for example, in various ways, for example by the position sensor recording the position of a closure mechanism in the valve, by the position sensor recording the flow rate through the valve, by the position sensor recording the actuating signal of a motor-driven closure mechanism or by similar mechanisms.

[0128] The drive control variable for the at least one delivery valve or the at least one discharge valve can be transmitted, for example, from the control device of the tire heat pressing apparatus to one of the mentioned valves in order to regulate the respective valve state. The drive control variable can be designed and configured as an analog and / or digital signal, for example, such as is customary for such control devices.

[0129] In order to fill the heat capsule with the respective process substance, a switching valve for hot gas at a relatively low pressure (for example between 2 and 8 bar), for example between 110 and 180 degrees Celsius, can be provided in the tire heat pressing apparatus, for example, and a further switching valve for delivering hot gas at a relatively high pressure (17 to 30 bar), for example 190 to 230 degrees Celsius. Then, for example, one or more further regulating valves can also be provided in the further delivery lines for the heat capsule in addition to the mentioned switching valves, by means of which the delivered gas flow or the pressure of the delivered gas flow can be regulated.

[0130] The extraction of process material located in an elastic heat bladder can be achieved, for example, by overpressure dominating the heat bladder and / or by extraction, for example, by means of a pump.

[0131] Furthermore, the method according to this specification and / or the tire hot pressing device according to this specification can also be designed and configured such that at least one of the at least one sensor value is provided by a heat bladder pressure sensor, which is designed and configured to measure the internal pressure in the heat bladder.

[0132] and / or

[0133] At least one of the sensor values ​​is provided by a delivery pressure sensor, which is designed and configured to measure the pressure in the delivery line for the heat bladder.

[0134] and / or

[0135] At least one of the sensor values ​​is provided by an exhaust pressure sensor, which is designed and configured to measure the pressure in the exhaust line for the heat bladder.

[0136] This pressure sensor can be any pressure sensor suitable for this pressure measurement and corresponding operating conditions in tire hot pressing equipment.

[0137] Here, such a pressure sensor can be any type of pressure sensor that is advantageously used for or applicable to the temperature and pressure relationship present in the corresponding tire hot pressing equipment. Within the scope of such equipment, for example, hot gases or liquids at relatively low pressures (e.g., between 2 and 8 bar) or relatively high pressures (17 to 30 bar) can be used (e.g., between 110 and 180 degrees Celsius and between 190 and 230 degrees Celsius). Therefore, for example, a pressure sensor suitable for temperatures up to 250°C and pressures up to 30 bar can be used.

[0138] Here, the pressure sensor can be placed at different locations within the medium supply device of the heat bladder and also within the heat bladder itself.

[0139] In a preferred design, for example, a pressure sensor can be provided in the delivery line, the discharge line, and the heat bladder itself. In this way, for example, a good picture of the condition of the heat bladder during filling and discharging, as well as during operation, can be detected, which enables the advantageous identification of potential problems with the equipment, the media supply device, and / or the heat bladder.

[0140] Furthermore, the method according to the present specification and / or the tire heat pressing apparatus according to the present specification can be designed and arranged such that at least one of the at least one sensor value is provided by a hot bladder temperature sensor, which is designed and arranged for measuring a temperature in the hot bladder,

[0141] and / or

[0142] at least one of the at least one sensor value is provided by a delivery temperature sensor, which is designed and arranged for measuring a temperature in a delivery line for the hot bladder,

[0143] and / or

[0144] at least one of the at least one sensor value is provided by a discharge temperature sensor, which is designed and arranged for measuring a temperature in a discharge line for the hot bladder.

[0145] The temperature sensors can also each be arranged in the same way, which can likewise be adapted to and arranged for the above-mentioned pressure and / or temperature relationships in the tire heat pressing apparatus.

[0146] In an advantageous design, the temperature sensors can also be arranged, for example, in the delivery line, the discharge line of the tire heat pressing apparatus and in the hot bladder, in order to be able to obtain as comprehensive a picture as possible of the state and the process in the context of the vulcanization process for vehicle tires.

[0147] Furthermore, it can be proposed that the ML model is designed and arranged as a neural network, which is trained according to the present specification at least with the aid of the sensor values and / or the control variables.

[0148] Here, for example, the ML model, the neural network and / or the training of the neural network can be designed and arranged according to the present specification.

[0149] In particular, the training of the neural network can be designed and set up as a so-called supervised learning, for example. Here, for example, so-called deep neural networks can be used. As a learning method, for example, a "deep learning" learning method can be used. Here, for example, the training data used to train the neural network can be designed and set up in such a way that one or more sensor values detected by one or more sensors and / or one or more control variables are associated with a state or characteristic variable of the tire heat pressing apparatus. The one or more sensor values detected and / or the one or more control variables can be detected, for example, at a certain point in time or also over a certain pre-set or pre-settable time period. This association of a state or characteristic variable with certain one or more sensor values is often referred to as a so-called "labeling" of the sensor data or control variables with the mentioned data. Furthermore, the one or more sensor data and / or the one or more control variables can also be associated with a characteristic variable, a quality characteristic variable or a description variable of a vehicle tire just present in the tire heat pressing apparatus at the point in time, or also with a characteristic variable or a state of the heat capsule.

[0150] In addition, supervised learning can also be designed and set up, for example, in such a way that the training data are formed in such a way that one or more of the above-mentioned characteristic variables or data are associated with a time series of sensor values of a certain sensor and / or a time series of values of a certain control variable. In a similar manner, it is also possible to associate a time series of sensor values originating from different sensors and / or a time series of values of different control variables with one or more of the above-mentioned characteristic variables.

[0151] In the case of an evaluation device comprising a simulation environment, for training the neural network according to the present specification at least also the sensor values of one or more soft sensors or virtual sensors simulated in the context of running a simulation program can be used, for example, according to the present specification, wherein the simulation environment is designed and set up for running a simulation program for simulating the tire heat pressing apparatus or parts or components of the tire heat pressing apparatus.

[0152] Also in the case of an evaluation device comprising a simulation environment with a simulation program for simulating the tire heat pressing apparatus or parts or components of the tire heat pressing apparatus, for training the neural network according to the present specification at least also the sensor values of one or more soft sensors or virtual sensors simulated in the context of running a simulation program can be used, for example, according to the present specification.

[0153] Here, the use of such soft sensors or virtual sensors values is similar to the use of values of real sensors of the tire heat pressing apparatus for training the neural network according to the present specification.

[0154] By virtue of training the respective neural network with one or more of the above-mentioned training data, the neural network can then be designed and set up in such a way that, on the basis of the data detected by one or more of the sensors of the tire heat pressing apparatus, the state of the tire heat pressing apparatus, of its components, for example the heat bladder, and of the vehicle tire currently present in the tire heat pressing apparatus can be derived. This can be designed and set up, for example, in such a way that the respective sensor values are used as input values for the neural network trained in this way, and the mentioned feature variables are then output values of the neural network or are derived from such output values accordingly.

[0155] In an alternative design, it is also possible, for example, to train a neural network with a so-called autoencoder structure by means of such sensor variables and / or control variables using unsupervised learning methods. With the trained autoencoder it is then possible, for example, to identify different states with respect to the tire heat pressing apparatus or the production process and / or the vehicle tire being processed. For example, on the basis of an evaluation of the identified states, it is then also possible, for example, to identify states associated with error situations. It is also possible to identify states associated with error situations or disturbances to occur in this way.

[0156] Thereby, for example, the effort for implementing the equipment and / or the structure for the process control and / or the facility monitoring of the tire heat pressing apparatus can be reduced, since the method can be designed and set up, for example, in such a way that only the sensors that are always present in the tire heat pressing apparatus are used in the context of the above-mentioned method, for example in order to achieve, for example, the normal operation of the vulcanization process.

[0157] In an advantageous design it can be proposed that the neural network comprises an autoencoder structure which is trained at least also with sensor values and / or control variables according to the present description by means of an unsupervised learning method.

[0158] Furthermore, the neural network can have a deep learning architecture and be trained at least also with sensor values and / or control variables according to the present description by means of a supervised learning method. The ML model can be designed and set up, for example, as a support vector machine, a gradient boosting tree model or a random forest model, which is set up at least also with sensor values and / or control variables according to the present description, respectively.

[0159] Here, the ML model, the support vector machine, the gradient boosting tree model and / or the random forest model can be designed and set up, for example, according to the present description. Here, for example, the setting of the mentioned ML model and also the preparation of the data as training data for setting the mentioned ML model can also be designed and set up according to the present description. Here, for example, the setting of such an ML model can be designed and set up by means of methods and processing means known for this purpose from the prior art.

[0160] The above object is also achieved by an evaluation device for monitoring a vulcanization process of a vehicle tire in a tire heat pressing apparatus, wherein the tire heat pressing apparatus is designed and configured for detecting at least one sensor value and / or at least one control variable of the tire heat pressing apparatus.

[0161] It is further proposed herein that the evaluation device comprises an ML model which is set by means of a machine learning method,

[0162] The evaluation device is designed and configured for receiving the at least one sensor value and / or the at least one control variable,

[0163] and the evaluation device and / or the ML model are further designed and configured such that, if an evaluation of the at least one sensor value and / or the at least one control variable by the evaluation device, in particular with the ML model, leads to:

[0164] - a malfunction or an anomaly of the tire heat pressing apparatus or of at least one part of the tire heat pressing apparatus,

[0165] - or a malfunction or an anomaly of the tire heat pressing apparatus or of at least one part of the tire heat pressing apparatus will occur in a foreseeable time,

[0166] a warning message is output.

[0167] Herein, for example, the evaluation device, the tire heat pressing apparatus, the vulcanization process of a vehicle tire, the sensor value, the control variable, the machine learning method, the ML model, the warning information, the malfunction, the anomaly or also the malfunction or anomaly which will occur in a foreseeable time, can be designed and configured in accordance with the present description.

[0168] As already explained in the context of the present description, when using the evaluation device according to the present description, the effort for implementing a facility monitoring and / or a process control for a tire heat pressing apparatus can be reduced in terms of devices and / or structures.

[0169] The tire heat pressing apparatus can comprise, for example, an elastic heat bladder which is designed and configured for supporting a vulcanization process of a vehicle tire located in the tire heat pressing apparatus.

[0170] Herein, for example, the elastic heat bladder can be designed and configured in accordance with the present description.

[0171] The evaluation device can be designed and configured for performing the method according to the present description, for example.

[0172] The above object is also achieved by a tire heat pressing apparatus for vulcanizing a vehicle tire,

[0173] The tire heat pressing apparatus comprises in particular an elastic heat bladder, which is designed and arranged for supporting a vulcanization process of a vehicle tire located in the tire heat pressing apparatus, wherein the tire heat pressing apparatus is designed and arranged for detecting at least one sensor value and / or at least one control variable of the tire heat pressing apparatus, and wherein the tire heat pressing apparatus further comprises an evaluation device according to the present specification.

[0174] Herein, the tire heat pressing apparatus, the elastic heat bladder, the vehicle tire, the at least one sensor value, the at least one control variable, and the evaluation device can be designed and arranged, for example, according to the present specification. BRIEF DESCRIPTION OF DRAWINGS

[0175] The application is explained below in more detail with exemplary reference to the enclosed drawings.

[0176] The drawings show:

[0177] Figure 1 An example of a tire heat pressing apparatus is shown;

[0178] Figure 2 A schematic view of a steam system for a tire heat pressing apparatus with associated control and evaluation device is shown;

[0179] Figure 3 A view of an alternative design of a steam system with associated control and evaluation device according to Figure 2

[0180] Figure 4 An exemplary design of a ML model configured as a neural network with an autoencoder structure is shown. DETAILED DESCRIPTION

[0181] Figure 1 An example of a tire heat pressing apparatus 100 according to the present specification is shown. Herein, for the sake of simplicity, only some of the main parts of such a tire heat pressing apparatus 100 (also referred to as tire heat press 100) are identified in Figure 1 Further details of such a tire heat pressing apparatus 100 are referred to the relevant professional literature.

[0182] The tire heat press 100 comprises two identical types of vulcanization stations for parallel vulcanization of a first tire blank 116 and a second tire blank 126. For vulcanization of the tire blanks 116, 126, the tire heat press 100 comprises two lower tire molds 112, 122 and two upper tire molds 110, 120, which surround the respective tire blank 116, 126 during the vulcanization process. Figure 1 ​In the middle, the upper tire mold 110, 120 and the lower tire mold 112, 122 are shown in an open state, which is used for loading the tire blanks 116, 126. During the vulcanization process of the tire blanks 116, 126, the upper tire mold 110, 120 is positioned on the lower tire mold 112, 122, respectively, and thus forms a closed state.

[0183] In order to support the vulcanization of the tire blanks 116, 126, the tire press 100 comprises two thermal capsules 114, 124, which are filled with hot steam at high pressure during the vulcanization process of the tire blanks 116, 126, expand within the respective tire blank 116, 126 and in this way press the tire blank against the lower tire mold 112, 122 or the upper tire mold 110, 120. In this way, in addition to the actual vulcanization of the tire material, for example also the tire contour and other embossed structures of the vehicle tire 116, 126 are imprinted into the tire blank 116, 126. Figure 1

[0184] The hot steam mentioned here for filling the thermal capsules 114, 124 is only one example of various gases or liquids that can be used in such a tire press 100. In the context of the vulcanization process, the gas or liquid can have a temperature of up to 100 degrees Celsius, advantageously up to 200 degrees Celsius and also advantageously up to 250 or 300 degrees Celsius.

[0185] Furthermore, the thermal capsules and the respective medium system can be designed and configured in such a way that, for example, air, steam, nitrogen and / or other gases, also having the above-mentioned temperatures, can be used as a gas. If, for example, a liquid is to be used instead, water is usually used.

[0186] Here, the thermal capsules 114, 124 can be designed and configured in such a way that they can withstand a pressure of up to 30 bar or more and a negative pressure of up to -1 bar at least inside the tire blank. The respective medium system can then also be designed for the pressure range.

[0187] Figure 2 The tire press 100 is shown with Figure 1 ​The water vapor medium system 102 of the right-hand side portion of the tire hot press 100 is shown. Here, a temperature sensor 450 is arranged in the region of the hot bladder 114, by means of which the temperature of the hot bladder 114 and / or the temperature of the interior of the hot bladder 114 can be determined. A first pressure sensor 410 is present at the output region of the hot bladder on the right-hand side of the medium system 102, which is configured to measure a pressure of between 0 and 4 bar. Furthermore, a second pressure sensor 420 is present, which is configured to measure a pressure of between -1 and 32 bar. By means of these two pressure sensors 410, 420 and the temperature sensor 450, the water vapor state in the hot bladder 114 can be determined with high accuracy in a wide range of pressures and temperatures.

[0188] The right-hand side region of the water vapor medium system 102 is designed and set up for the inflow of water vapor. An inflow channel 210 for the inflow of water vapor at a relatively low pressure of between 2 and 8 bar and a temperature of 110 to 180°C is present far left. In order to regulate the inflow in the first inflow line 210, for example, a switch valve 310 is provided in the line 210.

[0189] In the inflow channel 210, a temperature sensor 450 is arranged, by means of which the temperature of the water vapor in the inflow channel 210 can be determined. Figure 2 To the right of the first inflow line 210, a second inflow line 220 is provided for the inflow of water vapor with a high pressure of between 17 and 30 bar and a temperature of between 190 and 230°C. In the second inflow line 220, a further switch valve 320 is present for switching on and off the water vapor flow. After the two inflow lines 210, 220 have been combined, a regulating valve 322 is provided in the combined inflow line, by means of which the respective water vapor flow can be regulated more precisely. A safety valve 321 is provided downstream of the inflow line.

[0190] In the inflow channel 210, a temperature sensor 450 is arranged, by means of which the temperature of the water vapor in the inflow channel 210 can be determined. Figure 2 To the right of the high-pressure delivery line 220, a further delivery line 230 is arranged, which is set up for the inflow of further process gases or liquids. In the third inflow line 230, a further switch valve 335 is present for switching on and off the respective gas or liquid flow through the third delivery line 230.

[0191] All three delivery lines 210, 220, 230 then converge in a separate hot bladder delivery device 270, in which a main input valve 370 is provided.

[0192] The main outlet line 260 for discharging liquid or gas contained in the heat bladder 114 is located on the outlet side of the heat bladder. The aforementioned pressure sensors 410 and 420 are also located in the outlet line 260. The main outlet valve 360 ​​for regulating the discharge of medium from the heat bladder 114 is arranged downstream of the pressure sensors along the outlet direction. A third pressure sensor 430 for verifying that the medium is within a pressure range of 0 to 4 bar, and a fourth pressure sensor 440 for verifying that the medium is within a pressure range of -1 to 32 bar, are located downstream of the outlet valve 360 ​​in the outlet line 260. Another outlet line 240, equipped with a switching valve 340, is located downstream of the main outlet line 260.

[0193] A negative pressure line 250 with a switching valve 350 is located downstream of the main outlet line 260, exiting the outlet line 240. For example, gas or liquid contained in the heat bladder 114 can be actively drawn out via the negative pressure line. Here, for example, a suction line 450 or a pump located downstream of it (…) Figure 2 (Not shown) can be set to the maximum temperature of a medium in the range of -0.5 to -0.1 bar for building a vacuum and 60°C.

[0194] Figure 2 The valves 310, 320, 322, 324, 330, 335, 370, 340, 350, 360 and sensors 410, 420, 430, 440, 450 shown and / or explained above are examples of tire hot pressing equipment 100, which are necessary for or used in the preset vulcanization process of vehicle tires 116, 126 in tire hot pressing equipment 100. Figure 2 The sensors 410, 420, 430, 440, 450 and shown are... Figure 2 The sensors present in the valves 310, 320, 322, 324, 330, 335, 370, 340, 350, and 360 as necessary (e.g., for detecting the position or state of the respective valves 310, 320, 322, 324, 330, 335, 370, 340, 350, and 360) are, for example, examples of sensors 410, 420, 430, 440, and 450 of the tire hot pressing device 100, which are necessary for or used in the preset vulcanization process of the vehicle tires 116 and 126 in the tire hot pressing device 100.

[0195] In addition, Figure 2 The control device 130 is shown in the figure. The control device is designed and configured as a modular control device 130 having a central module 132, a first input / output module 134, and a second input / output module 136.

[0196] Here, the respective signal output lines run from the input / output modules 134, 136 to the different valves 310, 320, 322, 321, 330, 335, 370, 360, 340, 350 of the medium system 102. Control signals can be transmitted from the central module 132 of the control device 130 via the signal output lines to the aforementioned valves 310, 320, 322, 321, 330, 335, 370, 360, 340, 350 in order to thus adjust the respective valve positions.

[0197] Furthermore, respective signal input lines run from the temperature sensor 450 and the pressure sensors 410, 420, 430, 440 to the input / output modules 134, 136 in order to transmit the respective sensor values to the central module 132 of the control device 130.

[0198] This is symbolically represented in Figure 2 by the designations of the respective valves and sensors in the drawing with the output and input lines symbolically shown below the control device 130.

[0199] In the central module 132, a runtime environment for a respective control program for controlling the tire heat press 100 is provided. In order to cure the tire blank 116 introduced into the right-hand portion of the tire heat press 100, it is thus possible, for example, in the context of the flow of the control program, to control the inflow and outflow of the gases for the prescribed support of the curing process by means of the inputted sensor signals and outputted control signals for the respective valves 310, 320, 322, 324, 330, 335, 370, 360, 340, 350. In this way, the respective support of the curing process of the tire blank 116 is achieved by the water vapor respectively flowing into or out of the heat bladder 114.

[0200] Furthermore, in Figure 2 an edge device 500 is shown, which is an example of an evaluation device according to the present specification. Here, the edge device 500 comprises a neural network 502. Here, the neural network 502 is an embodiment of a ML model according to the present specification.

[0201] The edge device 500 is connected to the control device 130 via the field bus 139. For example, control commands for controlling the tire heat press 100 can be transmitted to the evaluation device via the field bus. Furthermore, position information of the valves 310, 320, 322, 324, 330, 335, 370, 360, 340, 350 of the medium system 102 can be transmitted to the edge device 500 via the field bus line 139. Furthermore, the measured values of the temperature sensor 450 and the pressure sensors 410, 420, 430, 440 can also be transmitted from the control device to the edge device via the field bus.

[0202] The neural network 502 is trained by means of a plurality of valve position values and temperature and pressure sensor values in such a way that the respective position values and sensor values are associated with one another, i.e. in the case of the sensor value or the respective detected sensor value combination whether the heat bladder is functioning properly, in the case of the respective sensor value or sensor value combination whether the heat bladder is faulty (and for example has a leak), or in the case of the sensor value or sensor value combination whether a failure of the heat bladder 114, 124 occurs within a foreseeable period of time. This period of time can for example be a period of time within which for example 10 vehicle tires, 50 vehicle tires or 100 vehicle tires are produced.

[0203] The valves 310, 320, 322, 324, 330, 335, 370, 360, 340, 350 of the medium system 102 and the sensors 410, 420, 430, 440, 450 of the medium system 102 are thus transmitted to the edge device 500 in the context of the control of the medium system 102 by means of the control device 130 during the production of the vehicle and are specified there as input variables in the trained neural network 502. If the neural network outputs information in the respective input data that the heating device 114 is functioning properly, the production continues without further notification.

[0204] If the neural network 502 outputs information after inputting the respective data, i.e. for example the data explained above, that a failure of the heat bladder 14 will occur within a foreseeable period of time, a corresponding warning message is output to the user. The warning message can for example be transmitted to the PC 600 via the data line 602 and output to the user via the PC 600. The warning message can for example comprise information about an expected failure of the heat bladder 114, 124 at a foreseeable point in time, wherein the foreseeable point in time in the message can also be specified in more detail.

[0205] If the neural network 502 outputs information after inputting the respective data, i.e. for example the data explained above, that a failure of the heat bladder 114 already exists, a corresponding warning message is output to the user, for example via the PC 600. Furthermore, it can be provided in this case that the respective message is also transmitted from the edge device 500 to the control device 130 via the field bus 139 and that a warning signal is output to the tire heat press 100 via the control device. The warning signal can for example be designed and configured as a red warning light and / or a corresponding audio signal. Furthermore, it can also be provided in this case that the respective parameters for controlling the medium delivery and discharge are changed in such a way that a quality of a quality or at least tolerable vehicle tire is also produced or can be produced at least for a certain period of time by means of the faulty heat bladder 114, 124.

[0206] Figure 3 It is shown Figure 2Another design possibility of the system shown in Figure 3 is shown in Figure 2 The medium system 102 with the hot capsules 114 shown in Figure 2 Furthermore, a control device 130 is shown with a central module 132 and input and output modules 134, 136, which control the input and output of the hot water vapor to the hot capsules 114 in the manner already explained in connection with Figure 3 The control device 130 shown in Figure 2 comprises an ML module 138, which comprises a neural network 502, which is arranged in an edge device 500 in the embodiment in

[0207] In the system shown in Figure 3 the neural network 502 is connected directly via a backplane bus of the control device 130 (not shown in Figure 3 ), via which the respective valve position values and sensor values are transmitted from the central module 132 of the control device 130 to the neural network 502 in the ML module 138. The respective information output by the neural network 502 is then in turn transmitted via the backplane bus to the central module 132 of the control device 130 and from there, if necessary, for example when a warning signal is generated, to the PC 600 via the data line 602. As already explained in connection with Figure 2 the respective user can then obtain via the PC 600 a respective warning about the status of the hot capsules 114, 124. The input and output signals of the neural network 502 and the control device 300 are trained in the context of the operation of the tire heat pressing device 100 correspond to this case explained in connection with Figure 2

[0208] Figure 4 An example of a design scheme of the neural network 502 shown in Figure 2 and 3 is shown in Figure 4 The neural network 502 shown in

[0209] This neural network 502 with an autoencoder structure is an example of an ML model according to the present description and can have been trained, for example, with training data according to the present description by means of unsupervised learning methods known to the person skilled in the art according to the present description.​

[0210] The autoencoder 502 has so-called nodes 510, which are structured in the example shown as five node layers 521, 522, 523, 524, and 525. Node layers 521 to 525 are... Figure 4 The nodes 510 are shown as overlapping nodes. Figure 4 Only a few nodes 510 are indicated with figure labels to simplify the display and make it clearer. The input data vector 560 (the vector is a one-dimensional matrix) has four data fields 561, 562, 563, and 564. Figure 4 As shown on the left, one of input fields 561, 562, 563, and 564 is connected to one of the nodes 510 of the first node layer 521 of the auto encoder 500. Data is input into the auto encoder 510 in this manner. The auto encoder 502 includes a so-called encoding region 530, which includes the first two node layers 521 and 522 of the auto encoder 500. Here, each node 510 of the first node layer 521 is connected to each node 510 of the second node layer 522.

[0211] Code region 540 is connected to encoding region 530, and the code region is composed of node layer 523. Here, each node 510 of the second layer 522 of encoding region 530 is connected to each node of code layer 523 of code region 540.

[0212] The auto encoder structure 102 has a decoding region 550 after the code region 540, which is composed of two node layers 524 and 525. The final node layer 525 is connected to the data fields 571, 572, 573, and 574 of the output data vector 570.

[0213] Now, for example, the autoencoder 502 can be trained such that an input data set 560 comprising, for example, position values of the valves 310, 320, 322, 324, 330, 335, 370, 360, 340, 350 and of the sensors 410, 420, 430, 440, 450 of the medium system 102 is input into the first node layer 521 of the encoding region 530 and then the parameters of the nodes 510 and node connections of the autoencoder 500 are adjusted via a learning method available for autoencoders or typical, such that the output data vector 570 output by the last node layer 525 of the decoding region 550 corresponds to the input data vector 560 or at least approximately corresponds to the input data vector 560. Such a typical learning method is, for example, the so-called "Back Propagation" method, "Conjugate Gradient Methods", the so-called "Restricted Bolzman Machine" mechanism or similar mechanisms or combinations thereof. For example, the neural network parameters to be determined during training can be weights of node connections or weights of input values of nodes ("Weights"), bias values of nodes ("Bias"), activation functions of network nodes or parameters of such activation functions (e.g. "Sigmoid function", "Logistic function", "Activation function"...) and / or activation thresholds of network nodes ("Threshold") or similar parameters.

[0214] The learning method of the autoencoder 500 shown in Figure 4 is an example of so-called "unsupervised learning".

[0215] For example, the neural network 502 according to the present specification can alternatively also comprise a network structure for supervised learning. For example, the network structures for supervised learning and unsupervised learning can also be combined. For example, the ML model 502 shown in the figures can comprise a neural network with an autoencoder structure, as it is shown, for example, in Figure 4 and / or also comprises a plurality of other network structures. Here, the autoencoder structure can relate to the number of the respective participating nodes and the dimension of the node layers, and also to the number of node layers, Figure 4 the examples of the autoencoder structure 502 shown deviate from this number.

[0216] Figure 4The autoencoder structure shown in Fig. 5 is an example of a so-called deep autoencoder 502, because not all nodes 510 of the autoencoder 502 are connected with the input or output of the autoencoder structure 502, so that there are so-called "hidden layers".

[0217] Generally, the autoencoder structure 502 can be built symmetrically to the code region, for example. Here, for example, the number of nodes 510 of each node layer 521, 522, 523, 524, 525 can decrease layer by layer from the input side up to the code region and then increase again layer by layer towards the output side. Thus, in this way, one layer 523 or multiple layers in the code region have the smallest number of nodes 510 in the context of the autoencoder structure 502. Figure 4 The autoencoder 502 shown in Fig. 5 is an example of the above-described symmetric autoencoder 502.

Claims

1. Method for monitoring a vulcanization process of a vehicle tire (116, 126) in a tire heat pressing apparatus (100), wherein detecting at least one sensor value and / or at least one control variable of the tire heat pressing apparatus (100), characterized in that at least one of the sensor values and / or at least one of the control variables is fed to an evaluation device, the evaluation device (138, 500) comprises an ML model (502) set by means of a machine learning method, and if an evaluation of at least one of the sensor values and / or at least one of the control variables by the evaluation device (138, 500) leads to the result that the tire heat pressing apparatus (100) or at least one part of the tire heat pressing apparatus (100) is faulty or abnormal, or that the tire heat pressing apparatus (100) or at least one part of the tire heat pressing apparatus (100) will be faulty or abnormal in a foreseeable time, the evaluation device (138, 500) and / or the ML model (502) outputs a warning message, all of the at least one sensor value are provided by sensors (410, 420, 430, 440, 450) of the tire heat pressing apparatus, which are necessary or used for a preset vulcanization process of a vehicle tire in the vehicle tire heat pressing apparatus, wherein the ML model (502) is designed and set as a neural network (502), which is trained at least also with the sensor values and / or control variables, or as a support vector machine, a gradient boosting tree model or a random forest model, which is designed at least also with the sensor values and / or control variables, respectively.

2. Method according to claim 1, characterized in that the tire heat pressing apparatus comprises an elastic heat bladder (114, 124), which is designed and set for supporting a vulcanization process of a vehicle tire (116, 126) located in the tire heat pressing apparatus (100).

3. Method according to claim 2, characterized in that the evaluation device (138, 500) and / or the ML model (502) are further designed and set such that, if an evaluation of at least one of the sensor values or at least one of the control variables by the evaluation device (138, 500), in particular with the ML model (502), leads to the result that the heat bladder (114, 124) is abnormal or faulty, or that the heat bladder (114, 124) will be abnormal or faulty in a foreseeable time, a warning message is output.

4. Method according to any one of claims 1 to 3, characterized in that the tire heat pressing apparatus (100) comprises a control device (130) for controlling the vulcanization process, and At least one of the at least one control variable comprises a drive variable output by the control device (130) within the scope of the vulcanization process, the drive variable being directed at the tire heat pressing apparatus (100) or a component of the tire heat pressing apparatus (100).

5. The method according to any one of claims 1 to 3, characterized in that the tire heat pressing apparatus (100) comprises a pressure sensor (410, 420, 430, 440) and at least one of the at least one sensor value is provided by the pressure sensor (410, 420, 430, 440), and / or the tire heat pressing apparatus comprises a temperature sensor (450) and at least one of the at least one sensor value is provided by the temperature sensor (450).

6. The method according to claim 2, characterized in that the tire heat pressing apparatus (100) comprises at least one delivery valve (310, 320, 322, 324, 330, 335, 370) for regulating the delivery of a gas or a liquid to the heat bladder (114, 124) and at least one discharge valve (340, 350, 360) for regulating the removal of the gas or the liquid from the heat bladder (114, 124), and at least one of the at least one sensor value is provided by a position sensor of at least one of the at least one delivery valve (310, 320, 322, 324, 330, 335, 370) and / or by a position sensor of at least one of the at least one discharge valve (340, 350, 360), and / or at least one of the at least one control variable comprises a drive variable for at least one of the at least one delivery valve (310, 320, 322, 324, 330, 335, 370) and / or for at least one of the at least one discharge valve (340, 350, 360).

7. The method according to claim 2, characterized in that at least one of the at least one sensor value is provided by a heat bladder pressure sensor (410, 420) designed and arranged for measuring an internal pressure in the heat bladder (114, 124), and / or at least one of the at least one sensor value is provided by a delivery pressure sensor designed and arranged for measuring a pressure in a delivery line (210, 220, 230, 270) for the heat bladder (114, 124), and / or at least one of the at least one sensor value is provided by a discharge pressure sensor (410, 420, 430, 440) designed and arranged for measuring a pressure in a discharge line (240, 250, 260) for the heat bladder (114, 124).

8. The method according to claim 2, characterized in that at least one of the at least one sensor value is provided by a hot bladder temperature sensor (450) designed and configured to measure a temperature in the hot bladder (114, 124), and / or at least one of the at least one sensor value is provided by a delivery temperature sensor designed and configured to measure a temperature in a delivery line (210, 220, 230, 270) for the hot bladder (114, 124), and / or at least one of the at least one sensor value is provided by a discharge temperature sensor designed and configured to measure a temperature in a discharge line (240, 250, 260) for the hot bladder (114, 124).

9. The method according to any one of claims 1 to 3, characterized in that the evaluation device (138, 500) comprises a simulation environment designed and configured to run a simulation program for simulating the tire heat pressing apparatus or for simulating parts or components of the tire heat pressing apparatus.

10. The method according to claim 9, characterized in that the evaluation device (138, 500) comprises a simulation environment having a simulation program for simulating the tire heat pressing apparatus or for simulating parts or components of the tire heat pressing apparatus.

11. The method according to claim 1, characterized in that the neural network (502) comprises an autoencoder structure trained at least also with the sensor values and / or control variables by means of an unsupervised learning method.

12. The method according to claim 1, characterized in that the neural network (502) has a deep learning architecture and is trained at least also with the sensor values and / or control variables by means of a supervised learning method.

13. An evaluation device (138, 500) for monitoring a vulcanization process of a vehicle tire (116, 126) in a tire heat pressing apparatus (100), wherein the tire heat pressing apparatus (100) being designed and configured to detect at least one sensor value and / or at least one control variable of the tire heat pressing apparatus (100), characterized in that the evaluation device (138, 500) comprises an ML model (502) set by means of a machine learning method, the evaluation device (138, 500) is designed and configured to receive at least one of the sensor values and / or at least one of the control variables, and the evaluation device (138, 500) and / or the ML model (502) are further designed and configured such that, if an evaluation of at least one of the sensor values and / or at least one of the control variables by the evaluation device (138, 500), in particular with the ML model (502), leads to a malfunction or an anomaly of the tire heat pressing apparatus (100) or of at least one part of the tire heat pressing apparatus (100), or at least a part of the tire heat pressing apparatus (100) will fail or behave abnormally in a foreseeable time, then an alert message is output, all of the at least one sensor value are provided by sensors (410, 420, 430, 440, 450) of the tire heat pressing apparatus, which are necessary or used for a preset curing process of a vehicle tire in the tire heat pressing apparatus, wherein the ML model (502) is designed and set up as a neural network (502) which is trained at least also with the sensor values and / or control variables, or the ML model (502) is designed and set up as a support vector machine, a gradient boosting tree model or a random forest model which are respectively designed at least also with the sensor values and / or control variables.

14. The evaluation device according to claim 13, characterized in that the tire heat pressing apparatus (100) comprises an elastic heat bladder (114, 124) which is designed and set up for supporting a curing process of a vehicle tire (116, 126) located in the tire heat pressing apparatus (100).

15. The evaluation device according to claim 13 or 14, characterized in that the evaluation device (138, 500) is designed and set up for carrying out the method according to any one of claims 1 to 12.

16. A tire heat pressing apparatus for curing a vehicle tire, wherein, the tire heat pressing apparatus (100) comprises an elastic heat bladder (114, 124) which is designed and set up for supporting a curing process of a vehicle tire (116, 126) located in the tire heat pressing apparatus (100), and wherein the tire heat pressing apparatus (100) is designed and set up for detecting at least one sensor value and / or at least one control variable of the tire heat pressing apparatus (100), characterized in that the tire heat pressing apparatus (100) comprises an evaluation device (138, 500) according to any one of claims 13 to 15.

Images