Methods for predicting the aging of sewer structures

Abstract

The invention relates to a method for predicting the aging of canal objects (2, 4, 6), which comprises the following steps: - Examination of a large number of channel objects (2, 4, 6) at two different times, whereby at least one characteristic value representing the technical condition of the respective channel object (2, 4, 6) is formed in each examination, - Determining an aging factor for each of the investigated canal objects (2, 4, 6) based on the determined characteristic values ​​and taking into account the time interval of the at least two investigations of the respective canal object (2, 4, 6), - Creating a generalized aging function (8) based on a multitude of identified aging factors, as well as a data processing system and a computer program product designed to carry out this procedure.

Description

[0001] The invention relates to a method for predicting the aging of canal objects in a wastewater canal, as well as a data processing system and a computer program product designed to carry out such a method.

[0002] Operating sewer networks, i.e., networks of wastewater drains, presents the economic and personnel challenge of efficiently planning, and especially proactively planning, the renewal, rehabilitation, and renovation of sewer infrastructure. To ensure that sewer infrastructure is only rehabilitated when truly necessary and can be maintained or even improved at a consistent level of quality over the long term, proactive planning is essential. This means it is desirable to understand how such sewer networks will age in the future in order to implement comprehensive planning.

[0003] Predictive models already exist that attempt to forecast the aging of such systems. These models currently attempt to make predictions for entire subnetworks consisting of numerous structures. Object-specific predictions concerning concrete individual canal components or structures are not possible with these models.

[0004] DE 10 2010 022 608 A1 discloses a method and robotic system for the internal inspection of a tubular object. To estimate the temporal development of damage, images or data recorded at different times are compared. DE 10 2012 022 489 A1 discloses a device for sewer inspection using a stereo camera. DE 100 48 826 A1 discloses a method and a device for recording age-related changes in technical systems such as electric motors, for which, based on a wear model and taking into account usage-dependent operating parameters, a remaining service life or probability of failure is calculated, for example.

[0005] The object of the invention is to provide an improved method for predicting the aging of canal objects, which makes it possible to make more precise predictions for individual canal objects or sections.

[0006] This problem is solved by a method having the features specified in claim 1, by a data processing system having the features specified in claim 13, and by a computer program product having the features specified in claim 14. Preferred embodiments are described below.

[0007] A key component of the method according to the invention is the creation of an improved model or function for predicting the aging or condition development of individual sewer components in a sewer network or pipe system, i.e., wastewater sewer components in a wastewater sewer network or pipe system. The method according to the invention is based on the examination of a plurality of sewer components, with each component being examined at at least two different times. In each of these examinations, at least one characteristic value representing the technical condition, i.e., the aging state of the respective sewer component, is generated. Such a characteristic value can be generated according to known standardized methods, which make it possible to express or define the overall condition, i.e., the overall aging state of the sewer component, preferably through one or more characteristic values.In each inspection, the degree of wear of the respective object is considered, and a characteristic value representing this is calculated. Preferably, the overall condition of the respective channel object is described by one or more characteristic values.

[0008] According to the invention, individual sewer components are considered instead of larger networks or subnetworks, and characteristic values ​​for the aging state of individual sewer components are determined. Such sewer components can be, for example, sewer structures, manholes, individual pipe sections, conduits, or even sections / partial areas of such structures, manholes, pipe sections, or conduits. Likewise, a subnetwork consisting of a multitude of sewer components can also be considered. In the field of wastewater sewer engineering, a pipe section is understood to be a section of sewer between two manholes.

[0009] Based on these determined parameters, aging factors are calculated for individual channel components. Such an aging factor can represent the aging process over a specific time interval, e.g., within one year. To calculate an aging factor, the two parameters determined at different times and the time interval between these two points in time are taken into account. In this way, the aging factor can reflect the change in condition over time, i.e., the change in the degree of wear, between these two points in time. Such an aging factor can be calculated, for example, according to the following formula: ALTFACTi=ZSTi,t2−ZSTi,t1t2−t1 where ZST i,t the state of the i-th channel object at time t.

[0010] Based on a multitude of aging factors determined in this way, a generalized aging function is created according to the invention. This aging function can then subsequently be used to predict the aging or changes in condition over time for specific channel objects in a channel system. The aging functions are thus based on the aging factors of individual channel objects, so that the average aging of a channel object can be determined and predicted.

[0011] Preferably, an aging prediction for one or more sewer components is subsequently performed based on the generalized aging function thus generated. These steps do not have to follow each other directly. The generation of the aging function can take place in a preliminary step, and the resulting aging function can then be applied repeatedly and independently of the system used to generate the aging function. The aging prediction can be based on an inspection of the sewer component, as described below. However, it is also possible to generate an aging prediction for one or more sewer components without examining them, for example, based on a characteristic value determined by other means. Instead of deriving this characteristic value from an examination, an average characteristic value of one or more sewer components can be used, for example.

[0012] The preferred method for generating the aging function is to examine canal objects of different ages at at least two different time points. This means that each canal object is examined at various times, and given the large number of canal objects, objects of different ages are examined at least twice. Aging factors for canal objects of different ages are then determined, and based on these aging functions, a generalized aging function is created that depends on the age of the canal objects and / or other factors. In other words, an aging function is created that represents the dependence of aging on the age already reached by a canal object. This allows for the subsequent application of the aging function, taking into account the age already reached by an individual canal object, to predict future aging or...Deterioration of the condition can be predicted. For example, the aging factors of many channel objects can form a data cloud depending on the age of the channel objects, from which a generalized aging function that takes age dependency into account can be derived. This then represents the average aging of a channel object of a specific age. This can be done, for example, as a further development of the formula mentioned above, based on the following formula: OLDFACTϕa=1|GRa|∑GRAALTFACTi with GR a = {i : ALT i =a}, where ALT i represents the age of the i-th attitude.

[0013] To create the aging forecast, an initial age and an initial aging state, or a characteristic value representing these, are preferably used for the individual channel object under consideration. Based on these, an aging forecast is created using the previously developed aging function. The initial age and the initial aging state can preferably be the current age and the current aging state, i.e., a characteristic value representing the current aging state.

[0014] The advantage of such a model or aging function is that object-specific aging predictions can be made, i.e., aging predictions for individual sewer components, and the change in the condition of the sewer component can be tracked at any given time. Based on the aging prediction for the individual components of a sewer system, a prediction can then also be created for a subnetwork or the entire sewer network.

[0015] The aging function, as described above, can be fitted, for example, by curve fitting of a multitude of aging factors, i.e., based on a data cloud formed by the aging factors. Such a fitting of the function can be particularly advantageous using an artificial intelligence system or machine learning (supervised machine learning or SML). Conventional methods, such as regression, can be used for this purpose. In such a function fitting, further data or metadata can be used, which may originate, for example, from inspection, master, and condition data of the sewer objects or a sewer system.Condition data can include, for example, the soil type surrounding the canal object, the type of wastewater conveyed, the traffic load above the canal object, geoinformation data, weather data, seismological data, or similar information.

[0016] Preferably, a multitude of aging factors are determined for different types, classes, and / or properties of duct objects, and aging functions are created for each of these duct objects or types or classes based on these factors. The individual duct objects or types of duct objects can be, for example, sections, specific components, and / or individual pipe sections or segments of a duct system. In the subsequent application of the aging function, a suitable aging function is selected for the respective duct object. Alternatively or additionally, it is possible to design the aging function in such a general or multidimensional way that it fits or maps a multitude of different duct objects or types, or supplementary data or condition data of duct objects, whereby preferably within the aging function...Within the aging function model, the type or nature of the channel object, as well as its other data or condition data as described above, can be taken into account during the prediction. A multidimensional aging function can be generated using multidimensional fitting methods during training.

[0017] According to another possible embodiment of the method, the various types, classes, and / or properties of sewer structures can take into account different geometries, dimensions, profiles, materials, uses, types of conveyed wastewater, and / or types of surrounding soil. This means that, preferably, individual or adapted aging functions, or multidimensional aging functions, are created for each of these different types, classes, and / or properties. For example, when predicting the aging of a specific sewer structure, a suitable aging function can be selected whose underlying type, class, and / or property corresponds to, or is as close as possible to, that of the sewer structure currently being assessed. Alternatively, these factors can be considered in the aging prediction via a multidimensional aging function.

[0018] The aforementioned characteristic values ​​of the sewer structures describe their technical condition, i.e., their state of aging or degree of wear, with the overall condition preferably being represented by such a characteristic value. These characteristic values ​​are preferably numerical and can be mathematically processed in the described procedure steps to calculate an aging factor and generate an aging function. Such characteristic values ​​can be determined using known standardized methods, for example, the SubKans method, the DWA standard DWA-M 149-3, the BFR wastewater guidelines, or another suitable method. These methods are known and allow for the generation of comparable characteristic values ​​representing the state of aging of a sewer pipe through categorized questions or analyses.

[0019] The aging prediction of a sewer system can be based on a past or previous inspection and / or rehabilitation of the system. In a specific variation of this method, the aging prediction can also be performed directly during or immediately after the inspection / rehabilitation. In the case of rehabilitation, the completed rehabilitation measures can be taken into account to generate an aging prediction for the rehabilitated sewer system. For example, inspection data can be transferred to an external, such as cloud-based, computer system. This system then uses the inspection data to calculate the necessary parameters, provides any additional data such as existing data, and performs the aging prediction based on a previously defined and stored aging function.Alternatively, the aging prediction could also be performed wholly or partially in a computing unit of an inspection device during the inspection of the sewer structure. For this purpose, the necessary data about the inspected sewer structure and / or the required aging functions can be pre-stored in the computing unit and / or retrieved from an external computer system via a data connection. Thus, according to one variant of the method, an aging prediction can be performed essentially in real time or near real time during the inspection.

[0020] The parameters for the technical condition of sewer structures are preferably determined by image analysis of pictures and / or videos taken or recorded within the sewer structure. This can preferably be done using automated image analysis. Such automated image analysis can utilize artificial intelligence modules that detect specific damage or aging conditions of a sewer pipe in the image and automatically determine or assign corresponding parameters. In this way, it is possible to very quickly analyze large quantities of images or video recordings and assign them the appropriate parameters. This parameter generation can be performed in real time or near real time during sewer pipe inspection using a sewer pipe inspection module, i.e., a camera module.Alternatively or additionally, a subsequent image evaluation is possible, in which older images and videos, which were taken at earlier times, can also be taken into account or evaluated.

[0021] Preferably, at least some of the characteristic values ​​are determined based on previously recorded examination data, preferably image data. This makes it possible to use archived examination data from the past to generate the aging model or aging function(s) as described above. It is essential that for each individual canal object used as the basis for this function or model development, at least two examinations are always carried out and evaluated at different times.

[0022] In addition to the method described above, the invention relates to a data processing system comprising means for executing a method according to the preceding description. This data processing system can comprise several, even independent, computer systems. In particular, there can be a first computer system used to determine or generate the described aging function. This is thus a modeling system. After the function has been generated or created, it can be applied in a second computer system for aging prediction. This is therefore an application system that can then be executed simultaneously on a multitude of computer systems.Furthermore, the invention relates to a computer program product comprising instructions that, when executed in a data processing system, cause the system to execute a method as described above. The computer program product can comprise two sub-products: a modeling program designed to generate the aging function on a modeling system as described above; and a computer program product in which the generated aging function is implemented, such that when this computer program product is executed on an application system, the previously generated aging function is used to generate aging predictions.

[0023] The invention is described below by way of example with reference to the accompanying figures. The figures show: Fig. 1 schematically a canal network and Fig. 2 a diagram with a multitude of identified aging factors and a constructed aging function.

[0024] The Fig. Figure 1 schematically shows a section of a sewer pipe network, i.e., a wastewater sewer network. The section shown has several manholes 2 and 4, with inlets provided at manholes 4. Manholes 2 and 4 represent sewer structures. Between manholes 2 and 4 are individual pipe sections 6. These pipe sections 6 and manholes 2 and 4 can be inspected using standard inspection systems to determine their state of aging. Based on established standards, a characteristic value can be calculated that describes or represents the current state of aging.

[0025] According to the invention, the individual sections 6 or channel objects 2, 4, 6 are inspected at two points in time, and a characteristic value representing the state of aging at each point in time is determined. Historical data can be used to determine the aging characteristic values ​​at different points in time. As described above, an aging factor is determined from the two characteristic values ​​determined for each of the individual channel objects 2, 4, 6.

[0026] The Fig.Figure 2 shows a large number of the determined aging factors 2 for sewer structures of varying ages plotted in a diagram. The aging factors are plotted against age as examples in the diagram. The aging factors represent the aging within a specific time interval, for example, within one year. An aging function 8 is created from the resulting data or point cloud. This aging function 8 can be achieved by fitting the numerous aging factors. In this way, a generalized aging function 8 is created, which represents the aging factor as a function of the current age of a sewer structure. Such generalized aging functions 8 are created for different types and categories of sewer structures, for example, for pipe sections 6 and manholes 2, 4, under various operating conditions.If a comparable canal object is inspected again, an aging prediction for that canal object 2, 4, 6 can be generated using a previously defined generalized aging function 8, given its current age. Based on the large amount of data, it is possible to generate aging predictions for individual sections 6 or other individual canal objects 2, 4, or sections of canal objects. Only an aging function suitable for the canal object needs to be selected, and preferably, the current age of the canal object must be known. Reference symbol list 2.4 shafts 6 postures 8 Aging function

Claims

[1] Method for predicting the aging of channel objects (2, 4, 6), which comprises the following steps: - Examination of a plurality of channel objects (2, 4, 6) at two different times, wherein at least one characteristic value representing the technical condition of the respective channel object (2, 4, 6) is formed during each examination, - determining an ageing factor for each of the examined channel objects (2, 4, 6) on the basis of the determined characteristic values ​​and taking into account the time interval between the at least two examinations of the respective channel object (2, 4, 6), - Creating a generalized aging function (8) based on a large number of determined aging factors. [2] Method according to claim 1, in which an aging prognosis of a channel object (2, 4, 6) is carried out on the basis of the determined aging function. [3] Method according to claim 1 or 2, in which channel objects of different ages are examined at at least two different times in order to determine aging factors for channel objects (2, 4, 6) of different ages and to create an aging function (8) as a function of the age of the channel objects (2, 4, 6). [4] Method according to one of the preceding claims, in which the aging function (8) is created by curve fitting of a plurality of aging factors. [5] A method according to claim 4, wherein the curve fitting is carried out using machine learning. [6] Method according to one of the preceding claims, in which a plurality of aging factors are determined for different types, classes and / or properties of channel objects (2, 4, 6) and, based on these, an aging function (8) is created for the different channel objects (2, 4, 6). [7] Method according to claim 6, wherein the different types, classes and / or properties of channel objects (2, 4, 6) take into account different geometries, dimensions, profiles, materials, uses, types of wastewater carried through and / or types of surrounding soil. [8] Method according to one of the preceding claims, in which the channel objects (2, 4, 6) are individual sections, components, sections and / or sections of sections (6) of a channel system. [9] Method according to one of the preceding claims, in which the characteristic values ​​for the technical condition are numerical values ​​which are determined, for example, according to the SubKans methodology. [10] Method according to one of the preceding claims, in which an ageing prognosis of a sewer object (2, 4, 6) is carried out on the basis of the determined ageing function (8) during or immediately after the inspection and / or rehabilitation of the sewer object (2, 4, 6). [11] Method according to one of the preceding claims, in which the characteristic values ​​for the technical condition of the channel objects (2, 4, 6) are determined by image evaluation of images and / or videos which have been recorded in the channel object (2, 4, 6), preferably by an automated image evaluation. [12] Method according to one of the preceding claims, in which the characteristic values ​​are formed on the basis of examination data recorded in the past. [13] A data processing system comprising means for carrying out a method according to any one of claims 1 to 12. [14] Computer program product comprising instructions which, when the computer program product is executed in a data processing system, cause the system to carry out a method according to one of claims 1 to 12.

Images