Method for predicting the air clearance in a vehicle's braking system, computer program product, control unit for a vehicle's braking system, and vehicle

Abstract

The invention relates to a method (200) for predicting brake clearance (LL), in particular wheel-specific brake clearance (LL), in a brake system (10) of a vehicle (100), comprising the steps of: • Acquiring (210) operating data (B) of the brake system (10), • Transmitting (220) the acquired operating data (B) to a control unit (110), • Determining (230) the last brake clearance (LL), in particular additionally storing (131) the values ​​in a database, • Processing (240) the operating data (B) by the control unit (110) using an algorithm for predicting brake clearance (LL), wherein the algorithm is based on historical and real-time data, and • Determining (250) a predicted brake clearance value based on the processed data. Furthermore, the invention relates to a computer program product and a control unit (110) for a brake system (10) of a vehicle (100), which are trained to carry out such a procedure.The invention also relates to a vehicle (100) which has such a control unit (110).

Description

The invention relates to a method for predicting air clearance, a computer program product, a control unit and a vehicle. In modern vehicles, the safety and efficiency of the braking system are of paramount importance. One crucial aspect is brake pad clearance, the distance between the brake pads and the brake disc when the brakes are not applied. This clearance significantly influences the efficiency and lifespan of the braking system. Insufficient clearance can lead to increased wear and overheating, while excessive clearance increases braking response time. Therefore, precise monitoring and prediction of the brake pad clearance are essential to ensure the safety and performance of the braking system. The prior art, as described, for example, in EP 2948690 A1, comprises a disc brake with clearance monitoring that enables continuous monitoring of the clearance. The solution disclosed therein uses a wear sensor and a brake control unit to monitor the clearance and adjust it as needed. The known prior art thus only allows for real-time determination of the clearance. DE 10 2013 100 786 A1 discloses a disc brake, in particular for a motor vehicle, comprising a clamping device, in particular with a brake lever, an adjusting device coupled to the clamping device, in particular with the brake lever, for adjusting for wear of brake pads and brake disc, a wear sensor for detecting a wear value of brake pads and brake disc, and a brake control unit. The disc brake has a clearance monitoring device with a control unit which is connected to the wear sensor and the brake control unit; and a corresponding method for monitoring a clearance of a disc brake. DE 10 2021 210 169 A1 describes a method for determining the current air clearance of a brake device of a motor vehicle, comprising the steps of capturing parameters to determine the start point and end point of a build-up process of a braking process using a sensor unit, determining the start point and end point of the build-up process based on the captured parameters using a processing unit, and determining the air clearance of the brake device based on the determined start point and end point of the build-up process using the processing unit. From DE 10 2024 112 835 A1, a method for determining the clearance of a vehicle brake is known, comprising the following steps: Providing a vehicle brake with a clearance to be determined between two friction partners; Determining and / or setting the initial position of the clearance between the friction partners; Adjusting, preferably manually, the clearance by rotating an adjusting element of an adjusting device of the vehicle brake to an end position at which the friction partners are in contact with each other; Determining an angle of rotation between the end position and the initial position; and Determining a clearance based on the angle of rotation at a known slope when rotating the adjusting element. It is therefore an object of the present invention to overcome at least one of the disadvantages described above, at least partially. In particular, it is an object of the invention to propose a method for predicting and / or estimating air clearance, and in particular to enable control of the braking system based thereon. The foregoing problem is solved by a method according to a first aspect of the present invention, by a computer program product according to a second aspect of the present invention, by a control unit according to a third aspect of the present invention, and by a vehicle, in particular a motor vehicle, according to a fourth aspect of the present invention. Further features and details of the invention will become apparent from the dependent claims, the description, and the drawings. Features and details described in connection with the method according to the invention naturally also apply in connection with the computer program product according to the invention and / or in connection with the control unit according to the invention and / or in connection with the vehicle according to the invention, and vice versa, so that the disclosure regarding the individual aspects of the invention always refers to each other or can refer to each other. According to a first aspect, the present invention relates to a method for predicting brake clearance, in particular wheel-specific brake clearance, in a vehicle's braking system, and preferably a braking system control system comprising the steps of: - acquiring operating data of the braking system, - transmitting the acquired operating data to a control unit, - determining the last brake clearance, in particular additionally storing the values ​​in a database, - processing the operating data by the control unit using an algorithm for predicting brake clearance, wherein the algorithm is based on historical data and real-time data, and - determining a predicted brake clearance value based on the processed data. The method according to the invention provides continuous monitoring, prediction, and / or estimation of the brake clearance in a braking system to enable optimal vehicle performance and safety. Historical and real-time data as well as algorithms are used for this purpose, allowing the brake clearance to be predicted and preferably the braking system to be adjusted or controlled accordingly. The procedural steps, processes, and procedures described here should not be interpreted as necessarily requiring their execution in the specific order discussed or presented, unless explicitly stated as a sequence. The procedural steps can occur, at least partially, simultaneously or sequentially, and the sequence is not limited to that defined by the numbering, allowing individual steps to be performed in different orders. It should also be understood that additional or alternative steps may be employed. The process begins with the acquisition of operating data from the braking system. This operating data encompasses a variety of parameters relevant to the condition and performance of the braking system. This data is acquired continuously or at specific intervals during vehicle operation to obtain up-to-date information and data on the braking system. The collected operating data is transmitted to a control unit. The control unit is designed to process and / or analyze the received data. Data transmission can occur via communication protocols and interfaces, in particular CAN bus and / or LIN bus. This enables fast and reliable data transmission. The control unit can be configured as a brake system control unit or as part of a more comprehensive vehicle control system, such as a driver assistance system. A further step involves determining the last brake pad clearance. Brake pad clearance refers to the distance between the brake pads and the brake disc when the brake is not applied. The determined brake pad clearance can also be stored in a database to save historical data for later analysis and / or comparison. This database can be part of the control unit or a separate system that communicates with the control unit. It can be advantageous to determine the brake clearance indirectly from the system behavior of the braking system. A pressure-volume characteristic curve (pV curve) is preferably used for this purpose. The pressure-volume characteristic curve (pV curve) in a vehicle's braking system describes the functional relationship between the volume of brake fluid introduced into the hydraulic system and the resulting pressure in the braking system. Using the pV curve, the necessary flow rate to generate or reduce the desired brake pressure can be determined. This enables a rapid and precise adjustment of the brake pressure to the respective driving situations. Furthermore, by recording the current pV curve during operation, influences such as temperature changes or component aging can be compensated for, thus achieving consistently high control accuracy. The moment the brake pad contacts the brake disc is called the friction point and marks the transition from volume build-up to force transmission. The current brake clearance can thus be indirectly determined, or at least inferred, from the position of this friction point in the pV curve. Furthermore, it is conceivable that at least one sensor is provided to record pressure or displacement measurements in order to detect the friction point. Here, the increase in brake pressure or a change in the displacement or force characteristic during braking can serve as an indicator of the contact point of the brake pads. This sensor data analysis can then be used for brake clearance monitoring and / or prediction. Current measured values ​​can be compared with reference values ​​and / or characteristic curves to identify deviations from the brake clearance. The control unit processes the operating data using an algorithm to predict brake clearance. This algorithm relies on historical and real-time data to enable accurate and reliable prediction of brake clearance. The algorithm may employ various data analysis methods to identify patterns and correlations within the data and predict brake clearance. Historical data includes previous measurements and records of brake clearance and other relevant parameters. Real-time data includes current measurements and states of the brake system and / or vehicle. By combining this data, the algorithm can predict future brake clearance. In a further step, a predicted clearance value is determined based on the processed data. This value indicates how the clearance is expected to behave in the (near) future. Continuous monitoring and prediction of the clearance allows the braking system to be optimally adjusted to ensure maximum braking performance and safety. Consequently, the prediction can be used to adapt, or regulate, the braking system, thus providing the driver with optimized braking behavior. Furthermore, it can be advantageous for the operating data to include at least one of the following parameters: brake pad wear, brake disc temperature, and brake pressure, particularly for a machine learning model to be created based on the processed operating data. As the brake pads wear, the brake clearance changes, which in turn affects braking performance and brake response. The brake disc temperature provides information about the thermal stress on the brake disc. High temperatures can affect the brake disc, which in turn can influence the brake clearance. Brake pressure is a measure of the force exerted on the brake pads and is a direct indicator of braking performance. A machine learning model can be preferably created based on the processed operational data to enable and / or support the prediction of brake pad clearance. This machine learning model preferably uses various data points as input, primarily, but not exclusively, those recorded since the last contact of the brake pads. These data points include, for example, but are not limited to, the turning radius and centrifugal forces or lateral accelerations, particularly during cornering. The centrifugal forces can be characterized by features of the lateral acceleration, such as maximum and integrated values. The time since the last brake application can influence the (de)compression of the brake pad after application and thus affects the brake pad clearance. Preferably, random variables are also implemented in the model to account for potential (unknown or unrecognized) influencing factors of real-world driving conditions. These random variables include, in particular, quantities that are not known or can only be determined imprecisely. Such influencing factors could include, for example, the timing and nature of the brake pad detachment from the brake disc, as well as the influence of vertical acceleration, such as that caused by cobblestones. Furthermore, it is conceivable that changes in pad thickness due to brake wear are also taken into account. It may be possible to use the maximum brake pressure during a brake application as an additional input, whereby the maximum brake pressure could, for example, trigger the pad wear adjustment mechanism. The applied machine learning model outputs the expected brake clearance and / or remaining dead volume. Furthermore, the determined brake clearance can be used in other learning processes as a predictive value for subsequent braking maneuvers. This allows for continuous improvement and adaptation of the model. Overall, considering the various data points and random variables enables the machine learning model to make a precise, or at least more precise, prediction of the brake clearance. Consequently, the safety and efficiency of the braking system can be increased. Preferably, the training of the machine learning model for predicting brake clearance in a braking system comprises the following steps: One step in training the model may include data collection. This data preferably includes historical and real-time operating data of the braking system, capturing various parameters such as brake pad wear, brake disc temperature, brake pressure, and other parameters and factors, particularly those described above. Data collection preferably occurs continuously during vehicle operation to obtain comprehensive and, in particular, current data on the braking system. The data can be acquired from sensors, control units, and other vehicle systems. The next step involves data processing. This step preferably includes cleaning (the raw data), feature extraction, transformation, and normalization of the data. Data processing can also include handling missing values, removing outliers, and scaling the data to ensure that all features are comparable. In a further step, feature extraction can be performed, in which relevant features are extracted from the raw data. These features can include, for example, the turning radius, centrifugal forces during cornering, the time since the last brake application, and random variables, i.e., single or multiple signal values ​​as well as their combination. Feature extraction serves in particular to identify the most important and informative features that contribute to predicting the airflow. It may be planned that a suitable machine learning model will be selected in a subsequent step. Various types of models suitable for this task can be used, e.g., neural networks. The selected model is trained primarily using the preprocessed and extracted data. Training can be performed in multiple iterations, with the model being adjusted in each iteration to improve accuracy. After training, the model can preferably be validated. Validation aims to ensure that the model is suitable and can make reliable predictions for new, unknown data. Finally, the model is applied accordingly. The trained model can be used to predict the brake clearance in real time and to take targeted measures to adapt and optimize the braking system. Applying the model enables continuous monitoring and prediction of the brake clearance to ensure the safety and efficiency of the braking system. Furthermore, it is conceivable that the operating data could be recorded during and / or after each braking process. This ensures that the data used to predict brake clearance is current and particularly relevant. Recording operating data during a braking process allows for the collection of real-time information about the condition and performance of the braking system. This data can provide insights into the behavior of the braking system under various operating conditions, such as different speeds, loads, and environmental conditions. In particular, recording data during braking allows for the detection of dynamic changes in the system that are relevant for predicting brake clearance. After braking, certain parameters of the braking system can change, such as the temperature of the brake disc or the condition of the brake pads. These changes can affect the brake clearance and are therefore relevant for more accurate prediction. Collecting data after braking also allows for the identification of long-term trends and patterns that can be important for predicting brake clearance. Furthermore, the prediction may include a confidence interval that quantifies the reliability of the prediction. A confidence interval is a statistical measure that indicates the range within which the true value of a parameter lies with a given probability. In the context of predicting the outcome of a lottery game, the confidence interval indicates how reliable the prediction is and how much the prediction might deviate from the actual values. Using a confidence interval provides both a single prediction and an assessment of the uncertainty. The machine learning model can evaluate this and make informed decisions based on this prediction. A larger confidence interval indicates greater uncertainty in the prediction, while a smaller confidence interval indicates higher reliability. Quantifying the reliability of the prediction using a confidence interval thus makes it possible to evaluate the accuracy and precision of the prediction. In particular, the use of a confidence interval allows the braking system to react to uncertainties and take appropriate action. It is also conceivable to determine the remaining dead volume of the brake fluid in the brake system, where the dead volume refers to the volume of brake fluid required to overcome the brake clearance. The dead volume can be influenced by various factors, such as brake pad wear, temperature changes in the brake system, and / or mechanical influences. By continuously monitoring and determining the dead volume, the brake system can be adjusted to ensure sufficient brake fluid is available to overcome the brake clearance. Preferably, the volume filled into the wheel brakes is determined based on the brake fluid delivery rate. This can be achieved, for example, by adjusting the plunger travel. The delivery volume is then the product of the plunger travel and the piston cross-section of the plunger.Alternatively, for other conveying systems, the conveying volume can be estimated, for example, from the conveying volume of an (ESC) pump. Furthermore, as explained above, the machine learning model can determine, output, or predict dead volume. It is also conceivable that the operational data could have a weighting that is particularly time-dependent. This time-dependent weighting allows for the consideration or exclusion of aging effects, thus improving the accuracy and reliability of the prediction. In this context, time-dependent weighting means that different weights are assigned to the operational data depending on their age. For example, current data could be weighted more heavily than older data, as it is likely to be more relevant for predicting the airflow or to allow for a more accurate prediction. Furthermore, it is conceivable that the weighting could be used to give greater weight to defined time periods or events that are particularly relevant for the prediction, such as (exceptionally) heavy braking. Optionally, the control of a pressure regulator in the brake system can be adjusted based on the anticipated clearance value to precisely regulate the brake pressure. The pressure regulator is a component of the brake system that regulates the brake pressure. By adjusting the pressure regulator's control, the brake pressure can be set to overcome the clearance and thus achieve effective braking. The pressure regulator's control can preferably be based on the anticipated clearance value, which is preferably predicted by a machine learning model. Regulating the brake pressure ensures that the brake pads are effectively applied to the brake disc, resulting in optimal braking performance. Furthermore, it may be possible to determine individual brake pressures for each wheel, taking into account individual brake clearances, to enable, for example, optimal braking performance. Individual brake pressure means that the brake pressure for each wheel is set separately, based on the specific conditions and requirements of that wheel. Since the conditions and requirements can differ for each wheel, depending on factors such as load, road surface conditions, and driving dynamics, determining the brake pressure individually for each wheel can allow for improved control of the braking system. Moreover, by considering individual brake clearances for each wheel, the brake pressure can be set precisely. Furthermore, it is conceivable that a volume estimate of the brake fluid in the braking system is performed. This volume estimate specifically involves determining the amount of brake fluid present or potentially present in the braking system, and / or the amount required for the system's operation. The brake fluid volume estimate can be used to determine the necessary volume to overcome the predicted air gap. Furthermore, it is conceivable that a pV characteristic curve is measured, particularly while stationary, for individual wheels and / or the entire system. This characteristic curve is measured especially during initial and repeated application to take into account the condition of a pre-compressed brake pad. The pressure-volume characteristic (pV characteristic) in a vehicle's braking system describes the functional relationship between the volume of brake fluid introduced into the hydraulic system and the resulting pressure within the braking system. Using the pV characteristic, the necessary flow rate to generate or reduce the desired brake pressure can be determined. This enables rapid and precise adjustment of the brake pressure to the respective driving situations. Furthermore, by recording the current pV characteristic during operation, influences such as temperature changes or component aging can be compensated for, thus ensuring consistently high control accuracy. Measuring the pV characteristic provides relevant data that can be used to predict the brake clearance. It can be advantageous to detect the transition from exceeding the brake clearance to pressure build-up in the wheel brake using at least one sensor, which in particular detects vibrations and / or acoustic signals. The transition from exceeding the brake clearance to pressure build-up marks the point in the braking process where the brake pads contact the brake disc and the build-up of braking torque begins. The at least one sensor can be, for example, an ultrasonic sensor, a vibration sensor, or an acoustic sensor. Ultrasonic waves measure the time it takes for the waves to be reflected by the brake disc and return. By measuring this time, the distance can be determined, and the transition from brake clearance to pressure build-up can be identified. Vibration sensors detect vibrations that occur when the brake pads contact the brake disc. These vibrations can be caused by mechanical interaction between the pads and the disc and indicate the beginning of pressure build-up. By detecting these vibrations, the transition can be recognized and the braking process controlled and / or regulated accordingly. Acoustic sensors, such as vibration sensors, detect noises generated during braking. These noises can be caused by the brake pads contacting the brake disc and serve as a further indicator of the transition from brake pad clearance to pressure build-up. By detecting these acoustic signals, the transition can be identified, and the braking process can be precisely controlled and / or regulated. Furthermore, it is conceivable that conditioning is performed when a defined limit for the air gap and / or the confidence interval is exceeded, in particular that the conditioning is model-based. The conditioning ensures, for example, that the air gap remains within defined limits and that the braking system functions effectively and reliably. The conditioning is preferably performed when the air gap exceeds a defined limit or when the confidence interval of the prediction becomes too large. Through conditioning, the air gap can be kept within acceptable limits, thus ensuring the reliable operation of the braking system. Model-based conditioning means that the adjustment of the airflow is based on a model, particularly a machine learning model, that describes the behavior of the braking system. This model can be based on historical data and real-time data, as previously described. Brake conditioning, in this context, refers to the process of adjusting or calibrating the brake system so that it operates within the desired parameters. Conditioning can be achieved through various measures. For example, a pressure pulse in the brake system can be used to reduce the brake clearance and ensure it remains within defined limits. Applying a pressure pulse of 1 to 2 bar for a short time can reduce the brake clearance. It is also conceivable that certain brake clearance sizes may hinder safe operation in the fallback position, as they result in an excessively large dead volume. In such cases, conditioning might involve setting a high / maximum system pressure while stationary to deliberately trigger the brake pad self-adjusting mechanism. This can be beneficial for maintaining driving safety. A warning for the driver and / or a passenger regarding excessive air clearance may also be provided. According to a second aspect, the present invention further relates to a computer program product comprising instructions which, when executed by a control unit, cause the control unit to execute a method according to one of the preceding claims. This provides the same advantages with respect to a computer program product according to the invention as have already been described with respect to a method according to the invention. In particular, the method can be a computer-implemented method. The computer program product can be implemented as computer-readable instruction code. Furthermore, the computer program product can be stored on a computer-readable storage medium such as a data disk, a removable drive, volatile or non-volatile memory, or an integrated memory / processor.Furthermore, the computer program product can be made available or provided on a network such as the internet, from which it can be downloaded or run online by a user as needed. The computer program product can be implemented using software, one or more special electronic circuits (i.e., in hardware), or in any hybrid form (i.e., using both software and hardware components). According to a third aspect, the present invention further relates to a control unit for a braking system of a vehicle, in particular a motor vehicle, wherein the control unit is configured to carry out the method according to the invention. This results in the same advantages with respect to a control unit according to the invention as have already been described with respect to a method and / or a computer program product according to the invention. The control unit can comprise at least one processor and / or at least one microprocessor. Furthermore, the control unit can be at least partially or completely integrated into a central control unit of the vehicle. However, it is also conceivable that the control unit is at least partially or completely integrated into one or more decentralized control units. The control unit is designed to perform the steps according to the invention, to process the corresponding signals from the sensors and / or other control units, and to transmit control commands to the various components of the braking system, thus enabling precise and efficient control of vehicle deceleration. The control unit can, for example, be designed as an electronic control unit connected to various sensors and / or actuators in the vehicle. These sensors can provide information about the vehicle's condition, driving conditions, and braking requirements, while the actuators make the necessary adjustments to the components of the braking system. According to a fourth aspect, the present invention further relates to a vehicle, in particular a motor vehicle, with a braking system comprising a control unit according to the invention, wherein the control unit is configured to execute the method according to one of the preceding claims. This provides the same advantages with respect to a vehicle according to the invention as have already been described with respect to a method and / or a computer program product and / or a control unit according to the invention. Further advantages, features, and details of the invention will become apparent from the following description, in which several exemplary embodiments of the invention are described in detail with reference to the drawings. The features mentioned in the claims and in the description can each be essential to the invention individually or in any combination. Figure 1 schematically shows a flowchart for a preferred embodiment of the method according to the invention, and Figure 2 shows a preferred embodiment of a vehicle according to the invention. The figures use identical reference numerals for the same technical features, even for different embodiments. The procedural steps, processes, and procedures described here should not be interpreted as necessarily requiring their execution in the specific order discussed or presented, unless explicitly stated as a sequence. The procedural steps can occur, at least partially, simultaneously or sequentially, and the sequence is not limited to that defined by the numbering, allowing individual steps to be performed in different orders. It should also be understood that additional or alternative steps may be employed. The method 200 for predicting brake clearance begins with the acquisition 210 of operating data from the brake system. This operating data comprises a variety of parameters relevant to the condition and performance of the brake system. This data is acquired continuously or at specific times during vehicle operation to obtain current information and data about the brake system. Preferably, the operating data includes at least one of the following data or parameters: brake pad wear, brake disc temperature, and brake pressure, and in particular, a machine learning model is created based on the processed operating data. As the brake pads wear, the brake clearance changes, which in turn affects the braking performance and the brake's response. The brake disc temperature provides information about the thermal load on the brake disc.High temperatures can affect the brake disc, which in turn can affect the brake pad clearance. Brake pressure is a measure of the force exerted on the brake pads and is a direct indicator of braking performance. The recorded operating data is transmitted to a control unit 220. The control unit is designed to process and / or analyze the received data. Data transmission can take place via communication protocols and interfaces, in particular a CAN bus and / or a LIN bus. This enables fast and reliable data transmission. The control unit can be designed as a brake system control unit or as part of a more comprehensive vehicle control system, e.g., a driver assistance system. The control unit 110 processes and analyzes the operating data B and determines the last brake clearance LL in step 230. The brake clearance refers to the distance between the brake pads and the brake disc when the brake is not applied. The last brake clearance LL is preferably stored in a database 231 to save historical data for later analysis and comparison.This historical data contributes to the accuracy of the prediction and takes aging effects into account. It can be advantageous to determine the brake clearance indirectly from the system behavior of the braking system. A pressure-volume characteristic curve (pV curve) is preferably used for this purpose. The pressure-volume characteristic curve (pV curve) in a vehicle's braking system describes the functional relationship between the volume of brake fluid introduced into the hydraulic system and the resulting pressure in the braking system. Using the pV curve, the required flow rate to generate or reduce the desired brake pressure can be determined. In the next step 240, the operating data B are further processed to determine the expected clearance value LS 250. For this purpose, an algorithm is preferably applied, the algorithm being based in particular on historical and real-time data to enable an accurate and reliable prediction of the clearance LS. The algorithm can include various data analysis methods to identify patterns and correlations in the data and to make a prediction of the clearance. Historical data includes previous measurements and records of the clearance and other relevant parameters. The real-time data includes current measurements and states of the brake system 10 and / or the vehicle 100. By combining this data, the algorithm can achieve a prediction of the future clearance. Here, a machine learning model based on the processed operating data B is preferably created. The machine learning model preferentially uses various data as input, which have been collected primarily, but not exclusively, since the last brake pad wear. This data includes, for example, but is not limited to, the turning radius and centrifugal forces. Lateral accelerations, especially during cornering, are a significant factor. These centrifugal forces can be characterized by features such as maximum and integrated values. The time elapsed since the last brake application can influence the (de)compression of the brake pad after application and thus affects the brake clearance. Preferably, random variables are also implemented in the model to account for potential (unknown or unrecognized) influencing factors of real-world driving conditions. These random variables include, in particular, quantities that are not known or can only be determined imprecisely. Such influencing factors could include, for example, the timing and nature of the brake pad detachment from the brake disc, as well as the influence of vertical acceleration, such as that caused by cobblestones. Furthermore, it is conceivable that changes in pad thickness due to brake wear are also taken into account. It may be possible to use the maximum brake pressure during a brake application as an additional input, whereby the maximum brake pressure could, for example, trigger the pad wear adjustment mechanism. The machine learning model used provides the expected brake pad clearance and / or remaining dead volume as output. The calculated brake pad clearance can also be used as a predictive value for subsequent braking maneuvers in further learning processes. This allows for continuous improvement and adaptation of the model. Additionally, random variables are introduced into the model to account for the variability and unpredictability of real-world driving conditions. These random variables include, in particular, parameters that are unknown or only imprecisely known. Examples include the timing and nature of brake pad separation from the disc, as well as the influence of vertical acceleration, such as that caused by cobblestones. In the long term, changes in pad thickness due to brake wear can also be considered.The maximum brake pressure during brake application is another input that triggers the pad wear adjustment mechanism and reflects the maximum compression of the brake pad. The applied machine learning model outputs the predicted brake clearance and / or remaining dead volume. The brake clearance determined during braking can then be used as a predictive value for subsequent braking maneuvers in further learning processes. This enables continuous improvement and adaptation of the model. By considering the various inputs and random variables, the model can make a precise or refined prediction of the brake clearance, thus ensuring the safety and efficiency of the braking system. Within the scope of the invention, it is conceivable that the operating data are recorded during and / or after each braking process. This ensures that the data used to predict the brake clearance is current and (particularly) relevant. Recording the operating data during a braking process makes it possible to gather real-time information about the condition and performance of the braking system. The expected clearance value LS is used to determine the remaining dead volume of the brake fluid. The dead volume is the volume of brake fluid required to overcome the clearance. By continuously monitoring and determining the dead volume, the brake system can be adjusted so that there is always sufficient brake fluid to overcome the clearance. Determining the dead volume can be done using sensors that monitor the brake fluid level in the system. In addition, as explained above, the machine learning model can determine and output a dead volume. Within the scope of the invention, it may be provided that the prediction includes a confidence interval that quantifies the reliability of the prediction. A confidence interval is a statistical measure that indicates the range within which the true value of a parameter lies with a certain probability. In the context of predicting the outcome of the lottery, the confidence interval indicates how reliable the prediction is and how much the prediction might deviate from the actual values. It is also conceivable that the operational data have a weighting that is particularly time-dependent. This time-dependent weighting allows for the consideration or exclusion of aging effects, thus improving the accuracy and reliability of the prediction. In this context, time-dependent weighting means that different weights are assigned to the operational data depending on their age. A further step is the regulation of the brake pressure 270 by a pressure regulator 11. The pressure regulator 11 is a component of the brake system that regulates the brake pressure. By adjusting the control of the pressure regulator 11, the brake pressure can be precisely set. The control of the pressure regulator can preferably be based on the expected clearance value, which is preferably predicted by the machine learning model. Regulating the brake pressure ensures that the brake pads are effectively applied to the brake disc and that optimal braking performance can be achieved. A volumetric estimation of the brake fluid 280 is performed to ensure that the brake system has sufficient brake fluid. Measuring a pV characteristic curve 290 allows the determination of the pressure-volume characteristic of the brake system. This characteristic curve helps to understand and optimize the behavior of the brake system under various operating conditions. It can be advantageous to detect the transition from exceeding the brake pad clearance to pressure build-up in the wheel brake using at least one sensor, whereby this sensor detects vibrations and / or acoustic signals. The transition from exceeding the brake pad clearance to pressure build-up marks the point in the braking process where the brake pads contact the brake disc and the actual braking process begins. This sensor could be, for example, an ultrasonic sensor, a vibration sensor, or an acoustic sensor. Finally, conditioning can be performed if certain limits for the air gap and / or the confidence interval are exceeded. This conditioning is preferably model-based and may include the application of a pressure pulse to reduce the air gap. Fig. 2 shows a vehicle 100 with a braking system 10 according to the invention, comprising a control unit 110 for carrying out the method according to the invention and at least one wheel brake 12. The control unit can comprise at least one processor and / or at least one microprocessor. Furthermore, the control unit can be at least partially or completely integrated into a central control unit of the vehicle. However, it is also conceivable that the control unit is at least partially or completely integrated into one or more decentralized control units. The control unit 110 is designed to perform the steps according to the invention, to process the corresponding signals from the sensors and / or other control units, and to transmit control commands to the various components of the braking system, so that precise and efficient control of the vehicle deceleration can be achieved. The control unit can, for example, be designed as an electronic control unit that is connected to various sensors and / or actuators in the vehicle. These sensors can provide information about the vehicle's condition, the driving conditions, and the braking requirements, while the actuators make the necessary adjustments to the components of the braking system. Furthermore, it can be advantageously provided that at least one pressure regulator 11 is included. This makes it possible to adjust the control of a pressure regulator 11 in the brake system 10 based on the anticipated clearance value in order to regulate the brake pressure (precisely). The pressure regulator 11 is a component of the brake system 10 that regulates the brake pressure. By adjusting the control of the pressure regulator 11, the brake pressure can be precisely set to overcome the clearance and thus achieve effective braking. The control of the pressure regulator 11 can preferably be based on the anticipated clearance value, which is preferably predicted by the machine learning model. Reference symbol list 10 Brake system 11 Pressure regulator 12 Wheel brake 100 (Motor) vehicle 110 Control unit 200 Procedure 210 Acquisition of operating data 220 Transmission of operating data 230 Determination of the last brake clearance 231 Storage in database 240 Processing of operating data 250 Determination of the expected brake clearance value 260 Determination of the remaining dead volume 270 Regulating brake pressure 280 Volume estimation of brake fluid 290 Measurement of a pV characteristic curve 300 Conditioning B Operating data LL last brake clearance LS expected brake clearance (prediction)

Claims

Method (200) for predicting brake clearance (BL), in particular wheel-specific brake clearance (BL), in a brake system (10) of a vehicle (100), comprising the steps: - Acquiring (210) operating data (B) of the brake system (10), - Transmitting (220) the acquired operating data (B) to a control unit (110), - Determining (230) the last brake clearance (BL), in particular additionally storing (231) the values ​​in a database, - Processing (240) the operating data (B) by the control unit (110) using an algorithm for predicting brake clearance (BL), wherein the algorithm is based on historical data and real-time data, and - Determining (250) a predicted brake clearance value (BL) based on the processed data. Method (200) according to claim 1, characterized in that the operating data (B) comprise at least one of the following parameters: brake pad wear, brake disc temperature and brake pressure, in particular that a machine learning model is created based on the processed operating data (B). Method (200) according to claim 1 or 2, characterized in that the operating data (B) are recorded during and / or after a respective braking process. Method (200) according to one of the preceding claims, characterized in that the prediction includes a confidence interval that quantifies the reliability of the prediction. Method (200) according to one of the preceding claims, characterized in that a remaining dead volume of the brake fluid of the brake system (10) is determined (260). Method (200) according to one of the preceding claims, characterized in that the operating data (B) have a weighting which is in particular time-dependent. Method (200) according to one of the preceding claims, characterized in that, depending on the expected clearance value (LS), the control of a pressure regulator (11) in the brake system (10) is adapted to regulate the brake pressure (270). Method (200) according to one of the preceding claims, characterized in that a wheel-individual brake pressure is determined taking into account a wheel-individual air gap (LL). Method (200) according to one of the preceding claims, characterized in that a volume estimation (280) of the brake fluid of the brake system (10) is carried out. Method (200) according to one of the preceding claims, characterized in that a measurement (290) of a pV characteristic curve, in particular while stationary, is carried out for individual wheels and / or the overall system. Method (200) according to one of the preceding claims, characterized in that the transition from exceeding the air gap (LL) to pressure build-up in the wheel brake is detected by means of at least one sensor, wherein the at least one sensor detects in particular vibrations and / or acoustic signals. Method (200) according to one of the preceding claims, characterized in that if a defined limit value for the airflow and / or the confidence interval is exceeded, conditioning (300) is carried out, in particular that the conditioning is model-based. Computer program product comprising instructions which, when executed by a control unit, cause the control unit to execute a method (200) according to one of the preceding claims. Control unit (110) for a brake system (10) of a vehicle (100), in particular a motor vehicle, wherein the control unit (110) is configured to carry out the method (200) according to one of claims 1 to 12. Vehicle (100) with a braking system (10) comprising a control unit (110) according to claim 14, wherein the control unit (110) is configured to perform the method (200) according to any one of claims 1 to 12.

Images