Bentley Systems (United States)

Abstract Crack detection has been an active research topic for civil infrastructure inspection. Over the last few years, many research efforts have focused on applying deep learning-based techniques...

Isolation valves are critical elements in water distribution system since they enable water utilities to perform maintenance on the system without the need to shut down large parts of the system. In this paper several isolation valve configurations in terms of the density of valves are used in a case study. The results show that the failure of an individual isolation valve can significantly impact any shutdown. The density of valves in the network determines impacts of inoperable valves during a shutdown. Several conclusions drawn from the critical valve analysis in this study could be applied to guide the isolation valve management. The key findings are that increasing the density of valves improves resilience and some valves are much more important than others.

This article shows that for pressure‐dependent demand (PDD) modeling, as pressures become very low, not only does the decrease in pressure affect outflow from orifices (faucets, showers, leaks), but it can leave some orifices above the hydraulic grade. The orifices above the hydraulic grade line would not have flowed at all. This has implications for the PDD function used in some hydraulic models in that a simple orifice equation cannot accurately describe such nodal demands. This article presents the results of lab experiments that measure the effect of very low pressures on demands. An equation that accounts for the influence of very low pressure and orifice elevation is proposed and verified. The impact of this equation on modeling low‐pressure situations is presented.

With increasing concerns about infrastructure sustainability, ubiquitous sensing is an integral part of smart infrastructure in the context of smart cities. It generates large data sets containing hidden patterns and intelligence, which must be effectively extracted to derive actionable wisdom to support decision-making. Thus, it is imperative to develop intelligent data analytics to extract intelligence from data. Various data analysis methods have been developed in recent decades, but the lack of robustness and data assimilation features prevents the previously developed methods from yielding adequately accurate results for time-variant data sets over a long duration. This paper proposes an improved deep belief network (DBN), a deep machine learning model, which is integrated with genetic algorithms (GAs) and the extended Kalman filter (EKF) for effective predictive modeling and efficient data assimilation. The proposed method uses a genetic algorithm to optimize the configuration of the DBN for the given problem. Then the DBN is trained in two steps, namely pretraining layer by layer and fine-tuning with either a conventional back propagation (BP) algorithm, namely BP-DBN, or an EKF that is generalized with a new algorithm for calculating the Jacobian matrix for many-layer DBNs, namely EKF-DBN, which was tested together with BP-DBN and a recurrence neural network (RNN) on three real cases with and without data assimilation. The comparison results showed that the EKF-DBN outperforms BP-DBN and RNN in both computational efficiency and accuracy for predictive modeling. In addition, EKF-DBN generates the error covariance matrix that enables the calculation of prediction confidence interval. This can be used to detect the anomalies in a real system.

Abstract This paper presents an economic analysis of pressure control solutions for leakage and pipe burst reduction. In detail, it explores the operating conditions under which the installation of ...

Continuous compaction control (CCC) systems utilize a mounted accelerometer, GPS unit, and data acquisition system on soil or asphalt compaction equipment, in order to collect data about the compaction process in real time and over 100% of the compaction evaluation area. During their use, CCC systems may collect anomalous data that can have an influence on reported compaction quality results, and which can lead to erroneous conclusions about the degree of compaction of the soil, or inappropriate regression models between CCC data and in situ measurement data. In the current study, an existing neighboring weighted-estimation method was applied to the results from a field study in which a variety of CCC and in situ test data had been collected, to quantify and filter out anomalous data. Other researchers have proposed that this approach can improve presented CCC data, resulting in improved regression relationships between CCC data and in situ measurement data. For the data set that was evaluated, the results indicate that the removal of anomalous data using this methodology did not yield a significant change in the relationships between the CCC and in situ test results.

While hydraulic models of water distribution systems are widely used for planning and design, there are other situations where such models can be helpful. One use case is that of emergency planning and response. Models can be used to simulate a wide variety of emergencies which can be used to prepare response plans. During an emergency, models can also be used in real time for decision support. This paper discusses some applications of models in these situations.

In most cases, water system design is based on a demand forecast at the end of some planning horizon based on the final configuration of the system at that time. This design approach (aimed at designing all the network at a time) is incompliant with its actual development, which instead takes place in phases. As a consequence, in order to follow the network demand and layout growth in time, practitioners prefer to sub-divide the whole construction life into various time phases thus including the different phases of construction in the network design.This work is aimed at analyzing and comparing three different phased approaches for constrained minimum-cost design of water distribution networks: the single-phase design with demand feedback, the multi-phase design without demand feedback and the multi-phase design with demand feedback. The difference between the single-phase design and the multi-phase design lies in the fact that whereas the former entails optimizing a single construction phase at a time, i.e. the current construction phase, the latter is based on the phasing of construction and then is aimed at optimizing the current construction phase and all the subsequent phases, included inside a certain temporal horizon, simultaneously. The demand feedback is here used as a pragmatic tool for updating the forecast at some specific time instant of the future demand growth: such an update is performed by setting the future demand growth equal to that really observed in the previous time phase. Alternatively, the predicted demand growth rate at the generic time instant can be kept equal to the value assumed at the time instant when the generic node appears, without taking account of the demand variation really observed in time in the node (absence of demand feedback).Applications to a real case study show that the multi-phase design with the demand feedback is the most reliable because it makes it possible to reduce the overall construction costs while attenuating the occurrence of pressure deficits in the various construction phases of the network. Optimal design for a single phase will virtually guarantee a sub-optimal solution over the long run.

A genetic algorithm (GA) model can be helpful in locating leaks and incorrectly closed valves in a water system in addition to calibrating for pipe roughness. This paper provides some practical suggestions to help users collect the right quality and quantity of data, manage the GA runs, and interpret the results. Optimization techniques work by adjusting demands, locating leakage hotspots, and changing pipe roughness or valve status so that the model matches the field observations for flow and pressure. However, all field data is inaccurate to some extent. The first key to success is that it is necessary to have head loss in the system that is significantly greater than the error in measurement. Otherwise, adjustments to the model are essentially random. In many water distribution systems, the head loss from the boundary nodes to the measurement points is small, and thus extra head loss must be generated. The easiest way to accomplish this is by opening one or more hydrants such that the flow increases by a known amount. This magnifies the discrepancies between the model and field data and makes the model much more sensitive to changes in parameters. Even with these high flows, it is important to measure the resulting pressure and the elevation of the pressure gauge to a high degree of accuracy (and remember that the elevation of the gauge is not necessarily the elevation of the model node).