Soft wet grounds such as mud, sand, or forest soils, are difficult to navigate because it is hard to predict the response of the yielding ground and energy lost in deformation. In this article, we address the control of quadruped robots' static gait in deep mud. We present and compare six controller versions with increasing complexity that use a combination of a creeping gait, a foot-substrate interaction detection, a model-based center of mass positioning, and a leg speed monitoring, along with their experimental validation in a tank filled with mud, and demonstrations in natural environments. We implement and test the controllers on a Go1 quadruped robot and also compare the performance to the commercially available dynamic gait controller of Go1. While the commercially available controller was only sporadically able to traverse in 12 cm deep mud with a 0.35 water/solid matter ratio for a short time, all proposed controllers successfully traversed the test ground while using up to 4.42 times less energy. The results of this article can be used to deploy quadruped robots on soft wet grounds, so far inaccessible to legged robots.

As metro lines continue to expand rapidly in urban areas, the excavation of twin tunnels in shallow depths using shield tunnelling methods has become widespread. By analysing field data obtained from an actual shield tunnelling project, it has been observed that the post-ground settlement occurring over the preceding tunnel during the excavation of the following tunnel in silty sand is approximately 42% of the green field settlement, which cannot be disregarded. Accurate approximation of the post-ground settlement is useful for preventing any damage due to excessive deformation and to determine the total ground settlement profile during twin tunnel construction stage. And yet, only a few number of studies have focused on investigating and predicting the postground settlement that occurs during twin tunnel construction in soft soils. Therefore, this study develops a transparent model using the multi-gene genetic programming (MGGP) method, enabling the prediction of postground settlement during twin tunnelling. Comparative analysis demonstrates that the proposed model is userfriendly and capable of generalising to unseen data. The reliability of the MGGP-based model has been validated through sensitivity and parametric analyses. Additionally, when estimating post-settlement during twin tunnelling, it is essential to consider the spacing between twin tunnels, soil cohesion, and crucial operational parameters of the shield, such as torque and face pressure.

This study aims to evaluate the effectiveness of the newly developed shield test device (VMS) by Tongji University in Shanghai for determining the muck state inside EPB shield chambers in fine-grained soils. Muck blockage inside shield chambers frequently results in significant stratum deformation and excessive damage to these machines. The engineering community requires real-time and effective monitoring technology to assess the muck flow state inside EPB shield chambers. Three motion modes of the muck composed of fine-grained soil near the partition have been identified: flow around the central axis, overall static and local adhesion. These modes correspond to the three characteristics of shear plate torque near the shield partition: the periodic fluctuation, the three-stage change of 'rise-fall-stability', and a noticeable amplitude difference at the top and lower sides of the chamber. Finally, a preliminarily validated wireless signalling-type shear plate device that can be assembled on the field shield and an actual measurement method was further developed to determine the muck state inside the shield chamber in real-time.

Large offshore wind turbines (OWTs) may encounter extreme misaligned wind and wave conditions throughout their lifetime, which could trigger side-to-side resonance of the OWTs and thus substantial reduction in fatigue life. This study aims to (1) understand the dynamic response of fixed-bottom OWTs under misaligned wind and wave conditions, and to (2) propose an active torque control algorithm for dynamic loading mitigation. For these purposes, an integrated aeroelastic model, coupled with an advanced soil-monopile interaction (i.e., p-y+M-theta model), is built in OpenFAST for the DTU 10MW OWT supported by a monopile in soft clay. The numerical results show that under misaligned wind and wave conditions, where the wave peak period is likely to approach the tower natural period, the dynamic loading along the side-to-side direction dominates the fatigue design of the OWT. To mitigate the side-to-side dynamic loading, an active torque control algorithm is designed with feedback from measured side-to-side tower vibration to enhance damping, as well as feedforward from measured incoming wave height to counteract external force. Through the use of the feedback-feedforward active torque controller, the side-toside dynamic loading of the 10 MW OWT is significantly reduced, with the fatigue life extended from 19 to 39 years.