Abstract: Systems and methods are disclosed for optimizing values of a set of tunable parameters of an autonomous driving vehicle (ADV). The controllers can be a linear quadratic regular, a “bicycle model,” a model-reference adaptive controller (MRAC) that reduces actuation latency in control subsystems such as steering, braking, and throttle, or other controller (“controllers”). An optimizer selects a set tunable parameters for the controllers. A task distribution system pairs each set of parameters with each of a plurality of simulated driving scenarios, and dispatches a task to the simulator to perform the simulation with the set of parameters. Each simulation is scored. A weighted score is generated from the simulation. The optimizer uses the weighted score as a target objective for a next iteration of the optimizer, for a fixed number of iterations. A physical real-world ADV is navigated using the optimized set of parameters for the controllers in the ADV.

Abstract: Systems and methods for generating labelled audio data and onboard validation of the labelled audio data utilizing an autonomous driving vehicle (ADV) while the ADV is operating within a driving environment are disclosed. The method includes recording a sound emitted by an object within the driving environment of the ADV, and converting the recorded sound into audio samples. The method further includes labelling the audio samples, and refining the labelled audio samples to produce refined labelled audio data. The refined labelled audio data is utilized to subsequently train a machine learning algorithm to recognize a sound source during autonomous driving of the ADV. The method further includes generating a performance profile of the refined labelled audio data based at least on the audio samples, a position of the object, and a relative direction of the object. The position of the object and the relative direction of the object are determined by a perception system of the ADV.

Abstract: According to some embodiments, described herein is a method and a system for guaranteeing safety at a control level of an ADV when at least a portion of a planned path generated by a planning module of the ADV is uncertain due to traffics and/or road condition changes. The planning module, when generating a path, also generate a confidence level of each segment of the path based on one or more of perception data, map information, or traffic rules. The confidence levels are decreasing further away from the ADV. When the control module of the ADV obtains the path and the associated confidence levels, the control module issue control commands to track only one or two segments whose confidence levels exceeds a threshold hold, and issue default control commands for the rest of the path.

Abstract: According to various embodiments, described herein is a method of creating a simulation environment with multiple simulation obstacle vehicles, each with a different human-like driving style. Training datasets with different driving styles can be collected from individual human drivers, and can be combined to generate mixed datasets, each mixed dataset including only data of a particular driving style. Multiple learning-based motion planner critics can be trained using the mixed datasets, and can be used to tune multiple motion planners. Each tuned motion planner can have a different human-like driving style, and can be installed in one of multiple simulation obstacle vehicles. The simulation obstacle vehicles with different human-like driving styles can be deployed to the simulation environment to make the simulation environment more resemble a real-world driving environment.

Abstract: An obstacle is detected based on sensor data obtained from a plurality of sensors of the ADV. Multiple trajectories of the obstacle are predicted with corresponding probabilities including a first predicted trajectory of the obstacle with a highest probability and a second predicted trajectory of the obstacle with a second highest probability. A cautionary trajectory of the ADV is planned based on at least one of a difference between the highest probability and the second highest probability or a consequence of the second trajectory. The ADV is to drive with a speed lower than a speed limit and prepare to stop in the cautionary trajectory. The ADV is controlled to drive according to the cautionary trajectory.

Abstract: An obstacle is detected based on sensor data obtained from a plurality of sensors of the ADV. A distribution of a plurality of positions of the obstacle at a point of time may be predicted. A range of positions of the plurality of positions of the obstacle may be determined based on a confidence level of the range. A modified shape with a modified length of the obstacle may be determined based on the range of positions of the obstacle. A trajectory of the ADV based on the modified shape with the modified length of the obstacle may be planned. The ADV may be controlled to drive according to the planned trajectory to drive safely to avoid a collision with the obstacle.

Abstract: In one embodiment, a computer-implemented method for optimizing a controller of an autonomous driving vehicle (ADV) includes obtaining several samples, each sample having a set of parameters, iteratively performing, until a predetermined condition is satisfied: determining, for each sample, a score according to a configuration of the controller based on the set of parameters of the sample, applying a machine learning model to the samples and corresponding scores to determine a mean function and a variance function, producing a new sample as a minimum of a function of the mean function and the variance function with respect to an input space of the set of parameters, adding the new sample to the several samples, and outputting the new sample as an optimal sample, where parameters of the optimal sample are utilized to configure the controller to autonomously drive the ADV.

Abstract: A user equipment (UE) receives a message that indicates a sidelink (SL) communication resource configuration to be used by the UE for communicating SL control information and SL data between the UE and another UE. The UE transmits SL control information according to the SL communication resource configuration, and transmits SL data according to the SL communication resource configuration. The SL control information and the SL data are transmitted by the UE without the UE receiving, in a downlink control information (DCI), a grant of communication resources.

Abstract: The invention provides a built-in micro-interface oxidation system for preparing terephthalic acid from p-xylene. The oxidation system includes a first reactor, a rectifying tower and a second reactor which are sequentially connected. A first outlet is disposed on a side wall of the first reactor; a first inlet is disposed on a side wall of the second reactor; a material inlet is disposed on a side wall of the rectifying tower; and a material outlet is disposed at a bottom of the rectifying tower. The first outlet is connected with the material inlet of the rectifying tower; the first inlet is connected with the material outlet of the rectifying tower. Micro-interface units are arranged in the first reactor and the second reactor for dispersing and crushing air into bubbles. Through disposing micro-interface units in reactors, problems of high energy consumption, high raw material consumption and low reaction efficiency are solved.

Abstract: Mechanisms are provided to implement a multi-layer analytics framework. The multi-layer analytics framework obtains a plurality of analytics from one or more analytics source computing systems, at least two analytics being written in different computer programming languages. The multi-layer analytics framework applies a wrapper to each of the analytics in the plurality of analytics to thereby generate wrapped analytics. The wrapper provides a unified interface for executing the analytics in the plurality of analytics regardless of the particular computer programming language used to create the analytics. The multi-layer analytics framework registers the wrapped analytics in an analytics registry, and executes an analytics pipeline comprising wrapped analytics in the analytics registry to perform an analytics operation based on the unified interface of the wrappers of the wrapped analytics.

Abstract: Described herein are a method of training a learning-based critic for tuning a rule-based motion planner of an autonomous driving vehicle, a method of tuning a motion planner using an automatic tuning framework that with the learning-based critic. The method includes receiving training data that incudes human driving trajectories and random trajectories derived from the human driving trajectories; training a learning-based critic using the training data; identifying a set of discrepant trajectories by comparing a first set of trajectories, and a second set of trajectories; and refining, at the neural network training platform, the learning-based critic based on the set of discrepant trajectories. The automatic tuning framework can remove human efforts in tedious parameter tuning, reduce tuning time, while retaining the physical and safety constraints of the ruled-based motion planner.

Abstract: In a wireless communication system, a grant-free uplink transmission is an uplink transmission sent from a user equipment (UE) to a base station that does not need a dynamic and explicit scheduling grant from the base station. Systems and methods are disclosed for performing hybrid automatic repeat request (HARQ) for grant-free uplink transmissions. Signaling relating to the ACK/NACK for the HARQ, as well as signaling relating to configuring the UE for grant-free uplink transmission, is also disclosed.

Abstract: Systems and methods for sidelink configuration are provided. A transmitting user equipment (UE) transmits configuration parameters relating to a sidelink transmission between the UE and a second UE via higher-layer signalling and then transmits the sidelink transmission to the second UE. The higher layer signalling is a layer above the sidelink transmission. The parameters may include feedback channel parameters or parameters for sidelink reference signal transmission by the second UE. The parameters may be selected by the transmitting UE, or they may be received initially from the network. The parameters may be part of a sidelink configured grant configuration (SL CG) signalled to the second UE. The sidelink transmission may be a configured-grant transmission to the second UE according to the SL CG configuration. The higher layer signalling may be PC5 signalling or RRC signaling.

Abstract: A treatment system and method for cephalosporin wastewater are disclosed. The treatment system includes: a flocculation and sedimentation device, an alkali reaction tank, a PAC reaction tank, a PAM reaction tank, a wastewater heat exchanger, a wastewater heater and an oxidation reactor that are connected with each other in sequence, wherein the wastewater heat exchanger is provided with a material inlet, a material outlet, a heat source inlet and a heat source outlet. An oxidized water from the oxidation reactor enters the wastewater heat exchanger from the heat source inlet, the heat source outlet is connected with a product canister, the product canister is connected with a membrane filtration device to realize concentration treatment of a landfill leachate, the material inlet is connected with the PAM reaction tank, and the material outlet is connected with the wastewater heater.

Abstract: A user equipment (UE) may determine that a transmission resource includes a first orthogonal frequency-division multiplexing (OFDM) symbol that is configured as a downlink symbol or as flexible, where the transmission resource is allocated for uplink (UL) transmissions during a time duration, and includes K transmission occasions (TOs). The UE may transmit a first UL transmission in the transmission resource omitting the first OFDM symbol. The first UL transmission includes K repetitions to be transmitted in the respective K TOs, and the K repetitions includes an initial transmission and at least one retransmission of the initial transmission.

Abstract: A built-in micro interfacial enhanced reaction system and process for PTA production with PX are provided. The system includes a reactor and a micro interfacial unit disposed inside reactor. The reactor includes a shell, an inner cylinder concentrically disposed inside shell, and a circulating heat exchange device partially disposed outside shell, inner cylinder having a bottom end connected to inner bottom surface of the shell in closed manner and an open top end, a region between shell and inner cylinder being first reaction zone, inner cylinder containing second reaction zone and third reaction zone from top to bottom, circulating heat exchange device being connected to inner cylinder and micro interfacial unit respectively. The invention can solve problems of large waste of reaction solvent acetic acid under high temperature and high pressure and being unable to take out the product TA in time during existing process of PTA production with PX.

Abstract: A method for training a machine learning model to generate a characteristic pattern, the method includes obtaining training data associated with a reference feature in a reference image. The training data includes (i) location data of each portion of the reference feature, and (ii) a presence value indicating whether the portion of the reference feature is located within a reference assist feature generated for the reference feature. The method includes training the machine learning model to predict a presence value based on the actual presence value in the training data. The predicted presence value indicates whether a portion of a feature (e.g., a skeleton point on a skeleton of a contour of the feature) is to be covered by an assist feature. The training is performed based on the training data such that a metric between a predicted presence value and the presence value is minimized.

Abstract: A method for determining a mask pattern and a method for training a machine learning model. The method for determining a mask pattern includes obtaining, via executing a model using a target pattern to be printed on a substrate as an input pattern, a post optical proximity correction (post-OPC) pattern; determining, based on the post-OPC pattern, a simulated pattern that will be printed on the substrate; and determining the mask pattern based on a difference between the simulated pattern and the target pattern. The determining of the mask pattern includes modifying, based on the difference, the input pattern inputted to the model such that the difference is reduced; and executing, using the modified input pattern, the model to generate a modified post-OPC pattern from which the mask pattern can be derived.

Abstract: A strengthening oxidation system of the external micro-interfacial unit for producing PTA with PX is provided, including: a reactor, a circulating heat exchange device and a micro-interfacial unit. The reactor includes an outer casing and an inner cylinder disposed concentrically inside the outer casing. The circulating heat exchange device is disposed at an exterior of the reactor, and is connected with the outer casing and the inner cylinder respectively, for regulating reaction temperatures of the first reaction zone, the second reaction zone and the third reaction zone inside the reactor in a reaction process of producing PTA with PX.

Abstract: A robust and accurate binary neural network, referred to as RA-BNN, is provided to simultaneously defend against adversarial noise injection and improve accuracy. Recently developed adversarial weight attack, a.k.a. bit-flip attack (BFA), has shown enormous success in compromising deep neural network (DNN) performance with an extremely small amount of model parameter perturbation. To defend against this threat, embodiments of RA-BNN adopt a complete binary neural network (BNN) to significantly improve DNN model robustness (defined as the number of bit-flips required to degrade the accuracy to as low as a random guess). To improve clean inference accuracy, a novel and efficient two-stage network growing method is proposed and referred to as early growth. Early growth selectively grows the channel size of each BNN layer based on channel-wise binary masks training with Gumbel-Sigmoid function.