Method for determining an adjusted current remaining value for an asset

Description

1. Technical area

[0001] The invention relates to a method for determining an adjusted current remaining time value for a plant. Furthermore, the invention relates to a computer program product and a corresponding determination unit. 2. State of the art

[0002] Residual value forecasts are well-known in the art. The residual time value in residual value forecasts is the time until the next switching point of a system. The residual time value can also be viewed as the remaining time, and the duration as the remaining time. The system can be any technical system, such as a traffic light system.

[0003] With a known and correct residual value forecast, the time until a technical system switches over can be calculated linearly over time. However, the residual value output by the forecast models can be significantly influenced by various factors, such as statistical uncertainties. Furthermore, unforeseen or unexpected events can occur, resulting in a new or revised residual value forecast that deviates from the previous forecast line.

[0004] As an example, the residual value forecast for a traffic signal system is presented below, in which the duration of a future switching point of a signal group is considered. With regard to the traffic signal system, for example, a public transport notification or a detector triggered by a road user can significantly influence the residual value.

[0005] The user of the residual value forecast expects a linear countdown of the remaining time value until the switchover point. For example, the user might be a driver who is on a direct route to the traffic light. It is therefore desirable to avoid fluctuations in the displayed remaining time that are perceived as random, thereby improving the credibility of the forecast system's behavior.

[0006] Traditionally, known approaches provide the user with the output of a neural model, such as the switching time and confidence. However, the prediction is simply hidden as soon as a predefined confidence value is exceeded.

[0007] A disadvantage of the known approaches, however, is that they inadequately consider the aforementioned factors, such as events. As a result, the remaining time forecasts have a low degree of reliability.

[0008] The document EP 3 340 204 A1 relates to a computer system for providing SPaT messages to a vehicle, wherein the SPaT messages include reliable timing parameters to influence the operation of the vehicle.

[0009] The present invention therefore has the objective technical task of providing a method for determining an adjusted current remaining time value for a plant, which is more reliable and efficient. 3. Summary of the invention

[0010] The above-mentioned object is achieved according to the invention by a method for determining an adjusted current remaining time value for a plant, comprising the steps: a. Providing a plurality of previous remaining time values, each with at least one associated confidence in a specific period, comprising b. at least one last previous remaining time value from the plurality of previous remaining time values ​​that immediately precedes the current remaining time value in the specific period; c. Receiving the current remaining time value with at least one current confidence; d. Determining the adjusted current remaining time value based on at least one previous remaining time value of the plurality of previous remaining time values ​​with the respective associated confidences depending on the current confidence; and e.Carrying out an action by the determining unit, wherein the action is selected from the group consisting of: outputting the adjusted current remaining time value and / or associated data on a display unit, storing the adjusted current remaining time value and / or associated data in a memory unit, and transmitting the adjusted current remaining time value and / or associated data to a computing unit.

[0011] Accordingly, the invention is directed to a method for determining an adjusted current remaining time value for a system. The remaining time value is the duration until the system's next switching time, as explained above. The remaining time value can also be considered the remaining time, and the duration the remaining time. In a traffic light system, such as a traffic light, the traffic light switches from one operating state to another at the switching time, for example, the traffic light changes from a green light to a red light. In other words, the signal change in this example occurs at the switching time.

[0012] In a first step, the input data set is received. The input data set can be received via one or more interfaces by the destination unit listed below, such as the processing unit.

[0013] The input dataset contains a plurality of previous remaining time values ​​with the respective confidence levels within a specific period. In other words, a time series with previous or known remaining time values ​​is considered. The time series therefore also has a first remaining time value and a last remaining time value within the period.

[0014] Furthermore, the input data set also includes a current remaining time value with its confidence. The current remaining time value is adjusted by the method according to the invention and provided as an adjusted current remaining time value.

[0015] Confidence is the certainty or probability that the deviation of the forecast remaining time from the actual or true remaining time does not exceed a specified value. 100% confidence means that the expected deviation of the forecast remaining time from the actual remaining time always lies within the specified deviation interval limits. With a confidence of 0%, no conclusions can be drawn about the expected deviation from the actual switchover time.

[0016] The current remaining time value is adjusted based on at least one previous remaining time value from the majority of remaining time values ​​with the associated confidence depending on the current confidence.

[0017] In this step, the last previous remaining time value can be used as the directly preceding remaining time value for adjustment. In addition, further history can also be used implicitly via the internal state, for example, when linearly extrapolating within the deadband. In this case, the starting value of this extrapolation can lie further in the past; it could possibly be the first value of the current cycle.

[0018] In other words, the history of previous remaining time values ​​and their confidence levels are evaluated depending on the confidence of the current remaining time. Consequently, all previous remaining time values ​​or just a portion of them can be used for adjustment.

[0019] The method according to the invention uses both the predicted remaining time and the confidence of the forecast from multiple time steps to smooth the current forecast. Consequently, a user advantageously perceives a smooth countdown of the current remaining time value, which can be displayed, for example, after the determination. The method can therefore also be referred to as a smoothing method.

[0020] For example, if a discrete event occurs and the forecast subsequently leaves a specified deadband, the adjusted current remaining time value is determined. Referring to the traffic signal system example above, the discrete event can be a public transport telegram input or a detector pulse. The determination can be made by replacing the continuous forecast with a combination of a linearly extrapolated previous forecast and the result of the updated forecast from a neural network.

[0021] In one embodiment, the remaining time values ​​and confidence levels are received from a computing unit or storage unit via at least one input interface. This embodiment has proven particularly advantageous with regard to efficient and reliable access to the data. The storage unit can be embodied as a volatile or non-volatile storage medium, such as a database or cloud storage. The computing unit can be embodied as any computing unit.

[0022] In a further embodiment, the remaining time values ​​and confidence levels are output values ​​of a neural network. Accordingly, the predicted remaining time and confidence of the remaining time forecast are outputs of a neural network, thus of machine learning.

[0023] In a further embodiment, the method further comprises the step Providing the current remaining time value as an adjusted current remaining time value if the at least one current confidence of the current remaining time value exceeds a specified maximum limit; determining the adjusted current remaining time value by extrapolating the last previous remaining time value if the at least one current confidence of the current remaining time value falls below a specified minimum limit, and providing the extrapolated remaining time value as an adjusted current remaining time value; or determining the adjusted current remaining time value by extrapolating the last previous remaining time value if the at least one current confidence of the current remaining time value falls below a certain percentage of the previous confidence of the last previous remaining time value and Provide the extrapolated remaining time value as an adjusted current remaining time value.

[0024] Accordingly, the current remaining time value is provided directly without adjustment if the confidence level is extremely high.

[0025] The last remaining time value with acceptable confidence is extrapolated linearly if the confidence is low.

[0026] The current remaining time value is calculated with acceptable confidence linearly extrapolated as long as the current remaining time value remains within a dead band around the extrapolated remaining time value, whereby the width of the dead band depends on the confidence of the current and previous forecast values. the linear extrapolation is corrected proportionately by the deviation from the current model output, whereby the proportion of the correction depends on the confidences of the forecast values.

[0027] In a further embodiment, extrapolation involves a counting unit linearly counting down a predetermined cycle time of the system. Accordingly, the current remaining time value can be counted down by seconds, minutes, or other cycle times, for example, of the traffic signal system. The counting unit enables reliable extrapolation.

[0028] In a further embodiment, the maximum limit value is a confidence greater than 90%, preferably 95%.

[0029] In a further embodiment, the minimum limit is a confidence between 20% and 30%.

[0030] In a further embodiment, the system is a traffic light or other traffic-related system. These limit values ​​have proven particularly advantageous with regard to the reliability of the adjusted current remaining time value as the output value.

[0031] One or more measures can be initiated after determining the adjusted current remaining time value as the output value of the method according to the invention. The measures can be implemented simultaneously, sequentially, or in stages.

[0032] First, the output value can be displayed to the user on a display unit of a computing unit. The forecast output is advantageously perceived by the user as a steady countdown, in contrast to the prior art. This improves confidence in the underlying forecasting system. Furthermore, the output value can be stored, and the output value itself or in the form of a corresponding message can be transmitted to another unit, such as a terminal device, a control unit, or other computing unit. The receiving computing unit can also initiate further appropriate actions upon receipt. Further actions include route planning, starting the engine of an autonomous vehicle, or other control measures.

[0033] The invention further relates to a determination unit. Accordingly, the method according to the invention is carried out by a determination unit. The determination unit is any computing unit. In addition to determining the adjusted current remaining time value, the determination unit can also initiate one or more of the above measures itself. This advantageously allows the measures to be implemented promptly and efficiently.

[0034] The invention further relates to a computer program product comprising a computer program which comprises means for carrying out the method described above when the computer program is executed on a program-controlled device.

[0035] A computer program product, such as a computer program means, can be provided or delivered, for example, as a storage medium, such as a memory card, USB stick, CD-ROM, DVD, or in the form of a downloadable file from a server in a network. This can be done, for example, in a wireless communications network by transmitting a corresponding file containing the computer program product or the computer program means. A program-controlled device can be, in particular, a control device, such as an industrial control PC or a programmable logic controller (PLC for short), or a microprocessor for a smart card or the like. 4. Brief description of the drawings

[0036] In the following detailed description, presently preferred embodiments of the invention are further described with reference to the following figures. FIG. 1 shows a flowchart of the method according to the invention for determining an adjusted current remaining time value for a system. FIG. 2 shows an exemplary remaining time forecast according to the prior art without smoothing. FIG. 3 shows an exemplary remaining time forecast according to an embodiment of the invention with smoothing. 5. Description of the preferred embodiments

[0037] In the following, preferred embodiments of the present invention are described with reference to the figures.

[0038] Figure 1 shows a flowchart of the method according to the invention with method steps S1 to S3. The individual method steps can be performed by the determination unit or its subunits. In the first two steps S1 and S2, the input data sets are received, previous forecasts, and the current forecast, as explained in detail above.

[0039] The forecasts can be output information from a neural network and include the forecast remaining time until the signal change as well as the confidence of the remaining time forecast.

[0040] According to the state of the art, only the current remaining time value of the remaining time forecast is output, regardless of the associated current confidence. The adverse effects are in Figure 2 These effects include, for example, 1) a drop in the forecast remaining time with low confidence followed by a recovery 2) jitter of the forecast with alternating rising and falling remaining time forecasts 3) short-term spikes with reduced confidence 4) strongly fluctuating remaining time forecasts, especially at the beginning of a phase

[0041] These effects are mitigated by the method according to the invention in step S3 as follows. In particular, the confidence of the current forecast is taken into account when determining the adjusted current remaining time value. Smoothing small fluctuations

[0042] The remaining time forecasts often exhibit minor fluctuations, which can lead to undesirable sign changes in the gradient and an uneven countdown. According to one embodiment of the invention, these fluctuations can be smoothed out by discarding the current value for small relative deviations from the previous forecast minus one second, e.g., less than 20% of the last value reduced by one, and adopting the previous value reduced by one as the forecast.

[0043] The pseudocode can be represented as follows: Smoothing larger fluctuations

[0044] If the fluctuations in the remaining time forecasts are larger than the small fluctuations mentioned above, this forecast can be smoothed according to one embodiment of the invention using a first-order IIR filter, where the filter coefficient depends on the confidence of the current forecast. This advantageously leads to a rapid adjustment of the forecast at high confidence, with only large steps being reduced, while at lower confidence the adjustment is correspondingly slower.

[0045] The pseudocode can be represented as follows: Integration of permanent small deviations of the smoothed forecast

[0046] Smoothing small fluctuations in the remaining time forecast can result in a permanently constant deviation in the remaining time forecast. To reduce such constant deviations, according to one embodiment of the invention, the deviation between the original and the smoothed forecast can be integrated, with this deviation being scaled using an integrator coefficient and the confidence of the current forecast value.

[0047] The pseudocode can be represented as follows: Smoothing spikes with reduced confidence

[0048] The forecast remaining time may experience short-term jumps with a simultaneous drop in confidence or even spikes. To compensate for these short-term disturbances, the forecast provided is e.g.below 80% of the confidence of the last forecast according to one embodiment of the invention and the last forecast is further counted down.

[0049] The pseudocode can be represented as follows: then current forecast = last forecast - 1 Smoothing forecasts with high or low absolute confidence

[0050] If a remaining time forecast has a particularly high confidence level, e.g., greater than 95%, with which the forecast can thus be considered reliable, then according to one embodiment of the invention, this forecast is not further smoothed but can be adopted directly. This forecast is already very close to the target value without smoothing.

[0051] The pseudocode can be represented as follows:

[0052] If a forecast has a very low absolute confidence, e.g. < 20%, and the last forecast has a higher confidence, then the last forecast remaining time is further counted down according to one embodiment of the invention.

[0053] Figure 3 shows the smoothed remaining time forecast according to one embodiment of the invention. The adverse effects can be reduced as follows: 1) The drop in the forecast remaining time is avoided. 2) The jitter of the remaining time forecast is smoothed. 3) Short-term spikes with reduced confidence are ignored. 4) Large fluctuations in the remaining time forecast are reduced.

[0054] The following pseudocode shows an exemplary implementation for the method according to an embodiment of the invention with smoothing in Python, taking into account additional constraints, e.g. no negative remaining time.

Claims

1. Computer-implemented method for determining an adapted current remaining time value for an installation, having the steps of: a. a determination unit receiving a plurality of previous remaining time values, each with at least one associated confidence, in a particular period of time via one or more interfaces (S1), having b. at least one last previous remaining time value from the plurality of previous remaining time values that immediately temporally precedes the current remaining time value in the particular period of time; c. the determination unit receiving the current remaining time value with at least one current confidence via the one or more interfaces (S2); d. the determination unit determining the adapted current remaining time value on the basis of at least one previous remaining time value of the plurality of previous remaining time values with the respective associated confidences on the basis of the current confidence (S3); and e. the determination unit carrying out a measure, wherein the measure is selected from the group consisting of: - outputting the adapted current remaining time value and / or associated data on a display unit, - storing the adapted current remaining time value and / or associated data in a storage unit, and - transmitting the adapted current remaining time value and / or associated data to a computing unit.

2. Method according to Claim 1, wherein the remaining time values and confidences are received by a computing unit or storage unit via at least one input interface.

3. Method according to Claim 1 or Claim 2, wherein the remaining time values and confidences are output values from a neural network.

4. Method according to one of the preceding claims, also comprising - providing the current remaining time value as the adapted current remaining time value if the at least one current confidence of the current remaining time value exceeds a defined maximum limit value; - determining the adapted current remaining value by extrapolating the last previous remaining time value if the at least one current confidence of the current remaining time value falls below a defined minimum limit value, and providing the extrapolated remaining time value as the adapted current remaining time value; or - determining the adapted current remaining time value by extrapolating the last previous remaining time value if the at least one current confidence of the current remaining time value falls below a particular percentage of the previous confidence of the last previous remaining time value, and providing the extrapolated remaining time value as the adapted current remaining time value.

5. Method according to Claim 4, wherein the extrapolation comprises linearly counting down a predetermined cycle time of the installation by way of a counting unit.

6. Method according to Claim 4 or Claim 5, wherein the maximum limit value is a confidence of greater than 90%, preferably 95%.

7. Method according to Claim 4 or Claim 5, wherein the minimum limit value is a confidence of between 20% and 30%.

8. Method according to one of the preceding claims, wherein the installation is a light signal installation or another installation in the traffic sector.

9. Determination unit for carrying out the method according to one of the preceding claims.

10. Computer program product having a computer program that has means for carrying out the method according to one of Claims 1 to 8 when the computer program is executed on a program-controlled device.