Method for processing sensor data from a working machine

Abstract

A method for processing sensor data from a working machine (2), comprising the steps: reading (S1) sensor data; determining (S2) a state of the working machine (2); specifying (S3) a filter depending on the determined state of the working machine (2); applying (S4) the specified filter to the read sensor data; and outputting (S5) the filtered sensor data to a functional device of the working machine (2), characterized in that if at least two states are determined when determining (S2), several filters are specified (S3), and the specified filters are used in combination when applying (S4).

Description

Technical field

[0001] The present invention relates to a method for processing sensor data from a machine. In this method, a filter is specified depending on a certain state of the machine. The specified filter is used to filter the sensor data. The present invention further relates to a control unit configured to execute such a method. The present invention also relates to a machine with such a control unit. State of the art

[0002] Methods and devices for filtering sensor data are known from the prior art. These methods use filters to remove unwanted parts of the sensor data. Fixed filters are used, with each filter being used in a specific application to filter out unwanted components of the sensor signals.

[0003] DE 10 2018 205 921 A1 relates to a method for generating an output signal, comprising the following steps: measuring a property with a first sensor, measuring the same property with a second sensor which has a larger measuring range than the first sensor, generating the output signal from a sensor signal of the first sensor, wherein during generation the sensor signal is influenced if, using the second sensor, it is detected that the measuring range of the first sensor has been exceeded.

[0004] DE 10 2019 220 118 A1 relates to a method for determining the direction of travel of a vehicle comprising the following steps: providing a first sensor for measuring a longitudinal acceleration of the vehicle and at least a second sensor for determining the rotational movement of a wheel of the vehicle; receiving an acceleration signal containing acceleration information from the first sensor; filtering the acceleration signal, resulting in a modified acceleration signal; determining the direction of travel of the vehicle based on the modified acceleration signal and based on the output signal of the second sensor. Description of the invention

[0005] The present invention relates, in a first aspect, to a method for processing sensor data from a working machine. The method can be carried out by a control unit of the working machine. The working machine can be a construction machine or an agricultural machine. Sensor data can comprise time-dependent data relating to signals such as acceleration, either linear or about a rotational axis, rotational speed, or torque of the working machine or of at least one element of the working machine. The control unit for carrying out the method can also be arranged outside the working machine.

[0006] The process involves reading sensor data. This sensor data, also called initial sensor data, can be acquired beforehand using a primary sensor. For example, a primary sensor could be an accelerometer, a gyroscope, a speed sensor, or a torque sensor. The process can include the acquisition of sensor data.

[0007] Furthermore, the procedure includes determining the state of the machine. For example, the state of the machine can be determined by a user, such as an operator, defining and specifying the state. Alternatively or additionally, the state can be determined by a machine itself. The state can consist of at least one working state and one movement state of the machine. For example, the working state can contain information about whether and what work the machine is performing at a given time. The movement state can contain information about whether and how the machine is moving.

[0008] Furthermore, the procedure includes specifying a filter depending on the particular state of the machine. This specification can involve determining the filter. A single filter can be specified, or alternatively, multiple filters can be specified. When specifying the filter, one can be selected from a plurality of predefined filters. A table can be used for this purpose, whereby a specific filter can be assigned to a particular state. The filter can be configured to filter out interference signals and, alternatively or additionally, interference frequencies. For example, the filter can be configured to filter out interference signals or interference frequencies from the input filter data.

[0009] Furthermore, the procedure includes applying the predefined filter to the input sensor data. Alternatively or additionally, several or all predefined filters are applied to the input sensor data. By applying the predefined filter to the input sensor data, the filter can remove parts of the sensor data. For example, the predefined filter can remove interference signals from the sensor data.

[0010] Furthermore, the method comprises outputting the filtered sensor data to a functional unit of the machine. This output can include sending a signal containing information about the filtered sensor data to the functional unit of the machine. The functional unit of the machine can be a second control unit of the machine. Alternatively or additionally, the second control unit of the machine can include such a functional unit. The second control unit can be connected, for example, electrically and electronically, to the first control unit, which can be configured to carry out the method according to the present invention. The second control unit can be configured to use the filtered sensor data. For example, a gradient calculation can be performed using the filtered sensor data by the functional unit and the second control unit.Alternatively or additionally, the filtered sensor data is output to a functional unit of the first control unit, which may be configured to execute the method of the present invention. In such a case, it is unnecessary to output the filtered sensor data from the first control unit.

[0011] This method allows a filter to be defined based on the specific state of the machine, and this defined filter can then be applied to sensor data. This enables the application of a specific filter to a particular state, eliminating the need for a single, dedicated filter to filter sensor data under different states. Using dedicated filters independent of the machine's state can result in certain interference frequencies, which the specific filter cannot remove, remaining in the sensor data. Furthermore, even if no interference signals or frequencies that the specific filter could remove are present in the sensor data, filtering with the specific filter can still lead to data manipulation. For example, a phase shift in the sensor data may occur.This method enables condition-dependent filtering of sensor data, filtering only when necessary and omitting it when not. Thus, during the operation of the machine, while the sensor data processing method is running, the filter(s) can be dynamically adjusted.

[0012] According to a further embodiment, the method can also include the reading of state data. The previously read sensor data from the first sensor can serve as state data. Alternatively or additionally, further, second sensor data can be read. For example, a second sensor can acquire second sensor data, which can be read as state data. The terms "first" and "second sensor" and "first" and "second sensor data" are used for nomenclature purposes only. If no further, second sensor data from a second sensor is read, patterns in the first sensor data from the first sensor can be recognized, and the state can then be determined based on the recognized patterns. Furthermore, the state of the machine can be determined based on the read state data. The state of the machine can be determined based on sensor data.For example, speed data can be read in as state data. The state of the machine can then be determined, for example, to be that the machine is moving at a certain speed in a certain direction.

[0013] Thus, the state of the machine can be determined based on read-in state data. This state data can be acquired via sensors and describes the physical state of the machine. Based on this determined state, the filter can be preset independently of the operator. This allows for automatic filter presetting or determination, independent of the operator, enabling the operator to concentrate on the further operation of the machine.

[0014] According to another embodiment, the filter can be a digital filter and can alternatively or additionally include a digital filter. A digital filter can be implemented via software on the control unit. Furthermore, when specifying the filter, a filter function and parameters of the filter function can be determined depending on the specific state of the machine. For example, a low-pass filter, a high-pass filter, a band-pass filter, or a band-stop filter can be defined as the filter function. For example, at least one parameter, such as a time constant, of one of the filters can be defined as a parameter.

[0015] This makes it possible to implement many different filters. You are not limited to a specific predefined filter, as is the case with analog filters. By implementing it as a digital filter, various filters can be dynamically used to process sensor data from the machine.

[0016] According to a further embodiment, if determining the state involves the machine traversing uneven ground, a low-pass filter can be specified when defining the filter. For example, when determining the state, it can be specified that the machine is traversing uneven ground by detecting a roll angle and, alternatively or additionally, a travel speed. When specifying the filter as a low-pass filter, at least one parameter, such as a time parameter, can also be specified and determined.

[0017] This method is therefore also applicable to machinery used on uneven surfaces. For example, sensor data from agricultural or construction machinery operated in fields or, alternatively or additionally, on construction sites can be processed. Such machinery might be equipped with sensors designed for on-road operation. This method allows the sensor data recorded by these sensors to be processed. This eliminates the need to equip such machinery specifically with off-road sensors for off-road use. By using the low-pass filter, consistently high-frequency signal components in the sensor data can be filtered out; these can occur, for example, when the machinery travels over uneven terrain.

[0018] According to a further embodiment, if the operation of a vibratory compactor of the machine is determined when defining the state, a bandpass filter can be specified when specifying the filter. The vibratory compactor can, for example, be a vibratory plate compactor, which can be arranged on the machine. Alternatively or additionally, other attachments, such as a belt conveyor, can be connected to the machine. For example, the state can be determined by providing a control signal, which can actuate the vibratory compactor, to the control unit, which can execute the method according to the invention. Thus, the state of the machine is easily determined. Alternatively or additionally, a bandpass filter can be specified instead of the bandpass filter. A table can be used to assign which filter should be specified for which piece of equipment.

[0019] This method allows for a response to interference frequencies or signals from a vibratory compressor, or more generally, from an attachment or implement of the machine, and the sensor data can be processed accordingly. Using the vibratory compressor's control signal, the operating state of the machine's vibratory compressor can be easily determined. The interference signals or frequencies from the vibratory compressor can be easily identified from a table, for example, or these interference frequencies may already be known. A suitable filter, perhaps defined in a table, can then be specified and applied to process the sensor data.

[0020] According to another embodiment, if determining the state of the machine indicates that the machine is operating, a bandstop filter can be specified. For example, the machine's operation can be inferred from a signal used to control the motor, or alternatively or additionally from a sensor within the motor. This sensor could be a vibration sensor on or in the motor. For instance, the machine's operation can be determined by identifying amplitude peaks in the vibration signal. Thus, the presence of initial vibrations and amplitude peaks indicates motor operation at a specific speed, while subsequent amplitude peaks and vibrations indicate motor operation at a different speed.Engine operation can be used to operate the drive motor of the machine for propulsion and, alternatively or additionally, for operating attachments such as a vibratory compactor. Engine operation can, for example, include at least one operating mode: forward travel, reverse travel, driving in a high gear, driving at high speed, and driving with torque. For instance, the bandstop filter can be specified using a table. Thus, if the engine operation is defined as high-speed operation, a first bandstop filter adapted to high-speed operation can be specified. If the engine operation is defined as low-speed operation, a second, different bandstop filter can be specified.The table allows for the specification and determination of a respective bandstop filter, whereby certain states can be assigned to specific interference frequencies and thus to specific filters.

[0021] This allows the method to incorporate an adaptive filter that adjusts to changing engine behavior and operation. Therefore, filtering the sensor signal is unnecessary when, for example, the engine is not running.

[0022] Furthermore, it is unnecessary for the filter, which is intended to filter out interference signals from the motor operation of the working machine, to be limited to a specific, fixed operating point of the motor.

[0023] According to a further embodiment, if at least two states are determined when defining the state, several filters can be specified, and the specified filters can be used in combination during application. For example, the specified filters can be used simultaneously, in parallel, alternatively, or additionally in a complementary manner. For example, a first state of the driven machine can be determined, in which the vibratory compressor is used. A second state of the driven machine can be determined, in which motor operation is determined. Then, a first bandstop filter for interference frequencies of the vibratory compressor and a second bandstop filter for interference frequencies of the motor operation can be specified. Both bandstop filters can then be used in combination during application, so that precisely the two bands corresponding to motor operation and the vibratory compressor are blocked.

[0024] This allows for the detection and determination of multiple states, such as two or more. Therefore, one or more filters can be specified, enabling responses to multiple states, for example, all specified states. The method is thus applicable even when more than one interference frequency is present.

[0025] According to a further embodiment, position information can be read from a sensor of the machine. For example, the position information can be read from the first sensor, and alternatively or additionally from the second sensor of the machine. Position information can be read once or multiple times from a sensor, for example, from the sensor's non-volatile memory. The position information can be hard-coded on the sensor's non-volatile memory. Alternatively or additionally, position information can be hard-coded on the first control unit and, for example, stored on the first control unit's non-volatile memory. The position information can then be read from the first control unit's memory containing the position information.The sensor can be configured to acquire sensor data and make it available for input. Furthermore, the position information can include details of the sensor's location within the machine, and the filter can be applied based on this position information. Position information can encompass the sensor's mounting location within the machine. For example, the first sensor for acquiring sensor data, such as rotational acceleration data, might be mounted in the front of the machine. The machine's motor might also be located in this area. Using this position information, the process can then define and apply the filter based on the position data.For example, interference signals or frequencies from the engine may be amplified at the front of the machine. Conversely, interference signals or frequencies from the implement, such as the vibratory compressor, may be amplified at the rear. Therefore, the position information can also reveal which interference signals or frequencies are most prevalent in the sensor data.

[0026] When specifying the filter, it is possible to omit the use of a filter for motor operation if, for example, the sensor is located far from the motor. However, if the sensor is located close to the motor or, for example, the vibratory compressor, a filter can be used and specified that is appropriate for the relevant interference signal or frequencies. Thus, the process for processing sensor data can also be adapted to the relative positioning of the sensor to the element emitting the interference signal or frequency, such as the motor or vibratory compressor.

[0027] According to a further embodiment, if a first filter is specified at a first time point and a second filter is specified at a second time point after the first, a transition filter can be specified depending on the first and second specified filters. Specifying the transition filter can include determining the transition filter. For example, the transition filter can be specified and determined by an average of the first and second filters. Furthermore, the filter with the transition filter can be applied for at least a certain period of time after the first filter has been applied and before the second filter has been applied. Alternatively or additionally, the second filter can be applied by first using the average value for the second filter.

[0028] This method allows for the provision and implementation of a transition between the first and second filters. The transition can be designed using the transition filter in such a way that the transition between the time period with the first filter and the time period with the second filter is smoothed in the filtered sensor data. Thus, post-filtering can be performed depending on a pre-filtering period.

[0029] A second aspect of the present invention relates to a control unit with input interfaces for reading sensor data. Furthermore, the control unit can include output interfaces for outputting filtered sensor data. The input interfaces can also be configured for reading status data. Furthermore, the control unit can be configured to execute a method according to an embodiment of the first aspect of the invention.

[0030] A third aspect of the present invention relates to a machine comprising a first sensor for acquiring sensor data and a first control unit according to an embodiment of the second aspect of the present invention. Furthermore, the machine may have a second control unit for forming a functional device. In addition to the first sensor for acquiring first sensor data, the machine may also have a second sensor for acquiring second sensor data, wherein the reading of status data can be performed depending on the sensor data acquired by the first or the second sensor. Filtered sensor data can be output to the second control unit, or alternatively or additionally to a memory or functional device of the first control unit. The machine may be a construction machine or an agricultural machine. Brief description of the characters Fig. Figure 1 schematically shows a working machine with a control unit according to embodiments of the invention. Fig. Figure 2 shows a method according to an embodiment of the invention, which is characterized by the Fig. 1 control unit shown in Fig. The work machine shown in section 1 is feasible. Detailed description of embodiments

[0031] Fig. Figure 1 schematically shows a machine 2 according to an embodiment of the invention. The machine 2 comprises a first control unit 4, a first sensor 8, a second control unit 10, and a second sensor 12. The first and second control units 4, 10 and the first and second sensors 8, 12 are electrically and electronically interconnected. Furthermore, the machine 2 comprises a vibratory compressor 6. The first control unit 4 has input interfaces 4a and an output interface 4b.

[0032] Fig. Figure 2 schematically shows the steps of a procedure for processing sensor data from a working machine 2. Fig.1. The first control unit 4 is configured to execute the steps of the procedure. Initial sensor data is read (S1). This initial sensor data is acquired by the first sensor 8 and transmitted to the first control unit 4 via the electrical and electronic connection through input interface 4a. Position information is then read (S6) from the first sensor 8. This position information, stored on non-volatile memory within the first sensor 8, is transmitted to the first control unit 4. Furthermore, status data from the machine 2 is read (S2.0). Additional, secondary sensor data, specifically status data, is acquired by the second sensor 12 and transmitted to the first control unit 4 via the electrical and electronic connection and input interface 4a. Based on this secondary sensor data, a status of the machine 2 is determined (S2).The second sensor data acquired by the second sensor 12 includes information on the roll angle of the machine 2. Furthermore, status data (S2.0) is read in via other control units of the machine 2 (not shown). For example, information is sent from a control unit of the vibratory compressor 6 to the first control unit 4. This information includes when and at what frequency the vibratory compressor 6 is operated. Additionally, information is sent from an engine control unit (not shown) of a motor of the machine 2 to the first control unit 4. This information includes when the motor is operated and in which mode it is operating, i.e., whether the motor is running at low or high speed.

[0033] The state S2 of the working machine 2 is determined based on the input state data. The state of the working machine 2 is defined as driving over an uneven surface if the roll angle information indicates this. Additionally, the operation of the vibratory compressor 6 of the working machine 2 is determined if the signal from the vibratory compressor 6's control unit indicates this. Furthermore, a specific motor operation of the working machine 2 is determined if the corresponding signal from the motor's control unit indicates this.

[0034] Furthermore, a filter S3 is specified. This involves determining S3.1 a filter function and S3.2 parameters of the filter function depending on the determination S2 of the state of the machine 2. Thus, a low-pass filter is specified when the machine 2 is traveling on an uneven surface. A corresponding bandstop filter is specified when the operation of the vibratory compressor 6 is determined as the state. Another bandstop filter is specified when a specific motor operation of the machine 2 is determined as the state. Additionally, a filter is specified depending on the position information. Since the first sensor 8 is located far away from the vibratory compressor 6 within the machine 2, the bandstop filter is specified with a weak filter function with respect to any interference signal from the vibratory compressor 6.The filter S3 can be specified via a table with entries on the states of the working machine 2 and on the respective filters.

[0035] If a first filter is determined at a first time point and a second filter is determined at a second time point after the first, then a transition filter S3 is specified depending on the first and second specified filters. For example, a filter is first determined with respect to an interference signal from the vibration compressor 6, and later a filter is specified and determined with respect to high-speed driving over uneven terrain.

[0036] The predefined filter S4 is applied to the acquired sensor data. This means that S4, the low-pass filter, and the corresponding band-stop filters are applied to the initial sensor data acquired in step S1. In this case, this initial sensor data consists of acceleration data from accelerometer 8. The band-stop filters and the low-pass filter are applied in combination to the signal to be filtered—the initial sensor data from sensor 8. This filters out the interference signals from engine operation, driving over uneven terrain, and the operation of the vibratory compressor 6 from the acquired sensor data.

[0037] Furthermore, the filtered sensor data is output S5. The filtered sensor data is transmitted from the first control unit 4 to the second control unit 10 via an electrical and electronic connection. The second control unit 10 is configured to use the filtered sensor data to control a function of the working machine 2 and thereby implement a functional device of the working machine 2. Reference sign 2 working machine 4 (first) control unit 4a Input interface 4b Output interface 6 vibratory compressors 8 (first) sensor 10 (second) control unit 12 (second) sensor S1 (Step) Reading sensor data S2.0 (Step) Reading status data S2 (Step) Determining a state of the working machine S3 (Step) Specifying a filter S3.1 (Step) Determining a filter function S3.2 (Step) Determining parameters of the filter function S3.3 (Step) Specifying a transition filter S4 (Step) Applying the specified filter to the read sensor data S5 (Step) Output of filtered sensor data S6 (Step) Reading position information for the (first) sensor

Claims

[1] Method for processing sensor data from a working machine (2), comprising the steps: reading (S1) sensor data; determining (S2) a state of the working machine (2); specifying (S3) a filter depending on the determined state of the working machine (2); applying (S4) the specified filter to the read-in sensor data; and outputting (S5) the filtered sensor data to a functional device of the working machine (2), characterized by , that if at least two states are determined when determining (S2) the state, several filters are specified (S3), and the specified filters are used in combination when applying (S4). [2] Method for processing sensor data from a working machine (2), comprising the steps: reading (S1) sensor data; determining (S2) a state of the working machine (2); specifying (S3) a filter depending on the determined state of the working machine (2); applying (S4) the specified filter to the read-in sensor data; and outputting (S5) the filtered sensor data to a functional device of the working machine (2), characterized by , that position information is read (S6) to a sensor (8; 12) of the working machine (2), wherein the sensor (8; 12) is configured to acquire the sensor data and make it available for reading (S1) the sensor data, wherein the position information contains information about the position of the sensor (8; 12) within the working machine (2), and wherein the filter is set (S3) depending on the position information read. [3] Method for processing sensor data from a working machine (2), comprising the steps: reading (S1) sensor data; determining (S2) a state of the working machine (2); specifying (S3) a filter depending on the determined state of the working machine (2); applying (S4) the specified filter to the read-in sensor data; and outputting (S5) the filtered sensor data to a functional device of the working machine (2), characterized by, that if a first filter is specified (S3) at a first time, and a second filter is specified (S3) at a second time after the first time, a transition filter is specified (S3.3) depending on the first specified filter and the second specified filter, and the filter is applied (S4) with the transition filter for at least a period of time after the first filter has been applied (S4) and before the second filter has been applied (S4). [4] Method according to one of the preceding claims, wherein state data is read in (S1; S2.0) and the state of the working machine (2) is determined (S2) depending on the state data read in. [5] Method according to one of the preceding claims, wherein the filter is a digital filter, and wherein when specifying (S3) the filter, a filter function is determined (S3.1) and parameters of the filter function are determined (S3.2) depending on the determined state of the working machine (2). [6] Method according to one of the preceding claims, wherein, when determining (S2) the state, driving on an uneven surface with the working machine (2) is determined, when specifying (S3) the filter, a low-pass filter is specified. [7] Method according to one of the preceding claims, wherein, when determining (S2) the state, an operation of a vibratory compressor (6) of the working machine (2) is determined, when specifying (S3) the filter, a bandstop filter is specified. [8] Method according to one of the preceding claims, wherein, when determining (S2) the state, a motor operation of the working machine (2) is determined, when specifying (S3) the filter, a bandstop filter is specified. [9] Control unit (4) with input interfaces (4a) for reading (S1) sensor data, wherein the control unit (4) is configured to execute a method according to one of the preceding claims. [10] Working machine (2) comprising a first sensor (8) for recording sensor data and a first control unit (4) according to the previous claim.

Images