39 results about "State distribution" patented technology

The invention relates to the field of unmanned aerial vehicle navigation and control, and particularly discloses an unmanned aerial vehicle three-dimensional track prediction method based on a pre-distribution state estimation model and particle swarm optimization, which comprises the following steps: initializing a particle set, and calculating an initial state mean value and a covariance matrix by adopting unscented Kalman filter (UKF); predicting the state of the particles by using unscented Kalman filter (UKF); sampling the particles based on the state distribution predicted by the UKF; particle weights are calculated and normalized; carrying out resampling on the particles; and calculating a three-dimensional track estimation value of the unmanned aerial vehicle and predicting a future track. According to the method for predicting the three-dimensional flight path of the unmanned aerial vehicle, high-precision state estimation of the UKF is introduced on the basis of particle filtering, so that the algorithm can effectively solve the problems of particle degradation and sample depletion encountered by a traditional PF algorithm in the three-dimensional flight path prediction of the unmanned aerial vehicle, and the prediction precision and stability are improved; and the future trajectory of the unmanned aerial vehicle can be accurately predicted in a complex three-dimensional environment.

The invention relates to the technical field of operation state evaluation, in particular to a switching power supply health degree evaluation method based on data analysis, which comprises the following steps of: establishing operation load correlation characteristics through mean value and range calculation, extracting continuous increasing paragraphs in a ratio sequence by using a difference mode, and calculating the health degree of a switching power supply; when the length of the interval exceeds a threshold value, a degradation starting point is recorded, a data segment is input into a long short-term memory network for time sequence modeling, offset is accumulated point by point in the modeling process and is compared with the threshold value in sequence, an abnormal path range is output, a degradation trend judgment result is generated, and a heat dissipation unit temperature rise and magnetic element jitter sequence is divided into four types of states; adjacent state changes are counted one by one, a proportion matrix is generated, convergence distribution is obtained through a Markov model through power accumulation, serious degradation and failure state distribution values are compared with a threshold value point by point, a critical interval is marked when the threshold value is exceeded, and life prediction distribution is formed; and life prediction can be combined with multi-dimensional state changes instead of depending on a single index.

The invention discloses an acceleration sensor output value correction method, and relates to the technical field of data processing, and the method comprises the steps: S1, analyzing the data drift trend of historical operation big data, and carrying out the clustering division according to the drift fault type of a known acceleration sensor; s2, on the basis of fault development type state distribution of a plurality of known acceleration sensors, obtaining a fault development type pointed by real-time operation data of a target acceleration sensor, and marking support factors in the fault development type pointed by the real-time operation data of the target acceleration sensor, determining that the real-time operation data of the target acceleration sensor points to the fault development type feature data set; and S3, predicting the hardware drift state of each stage of the fault development type of a plurality of known acceleration sensors, generating the hardware drift state of the stage to which the real-time operation data of the target acceleration sensor points, marking a hardware drift interval, and generating an output value correction value of the acceleration sensor.

The invention provides an agent inference method and a related device. The method comprises the steps of obtaining information of a robot and an environment, and constructing a generation model of an intelligent agent according to the information of the robot and the environment; the intelligent agent interacts with the environment through the generative model, the generative model deduces the real state and outputs the prior state distribution, and the generative model is adjusted according to the prior state distribution to minimize the surprise degree; wherein the free energy is obtained according to the expected value of the logarithm of the KL divergence of the variation distribution and the prior state distribution and the quotient of the variation probability and the conditional probability, and the free energy is minimized to minimize the surprise degree; the intelligent agent has more efficient perception ability in an incomplete information environment, the model training efficiency is improved, environment interaction develops towards an interested state, and falling into a local optimal solution is avoided.

The application provides a network element state management method and device, electronic equipment and a storage medium, wherein the method comprises: decoupling a service area in a to-be-processed satellite network from a preset satellite, and defining the service area as a geographical location area, wherein the service area in the to-be-processed satellite network is coupled with the preset satellite; determining a mobile network location state of the to-be-processed satellite network based on the geographical location area; hosting other states of the to-be-processed satellite network to a local user equipment to obtain a distributed local state library; running a state management process of key signaling of the to-be-processed satellite network based on the distributed local state library; and obtaining a stateless space-ground integrated mobile network based on the mobile network location state, the distributed local state library and the state management process, wherein the stateless space-ground integrated mobile network is a network composed of satellites and ground networks without state migration. The application can avoid signaling storm.

The invention discloses a scheduling method for real-time frequency regulation and control and distributed unit collaborative optimization of a power grid, and the method comprises the steps: calculating a power distribution proportion through employing a dynamic programming algorithm according to the charge state distribution of unit clusters, and determining a priority response sequence of each cluster; transmitting the preliminary allocation instruction to a corresponding cluster through a communication network, and correcting a transmission delay influence by adopting a feedback mechanism to obtain an adjusted power output scheme; according to the adjusted power output scheme, if it is judged that a local fault signal exists, switching to power distribution of a standby cluster is carried out, and a final batch coordination instruction is determined; extracting an execution parameter from the final batch coordination instruction, and testing a response effect by adopting a simulation verification module to obtain an optimized scheduling strategy; and generating a real-time monitoring report through the balanced resource use data, and determining whether the system stability index meets the requirement or not.

The invention relates to the technical field of remote scheduling, in particular to a remote scheduling method and system for roadbed and pavement construction equipment, and the method comprises the following steps: carrying out the division based on task units, recognizing an equipment track inflection point and an operation state conversion sudden change interval, analyzing the equipment operation direction consistency and operation state distribution, and evaluating the linkage task scheduling efficiency. And analyzing a stress distribution trend of the construction area, calibrating an abnormal level, identifying a corresponding number, and outputting an intelligent collaborative scheduling instruction set of the construction equipment. According to the method, by accurately extracting and analyzing the running track and the operation state of the equipment, the abnormal interval and the deviation area can be effectively recognized, resource allocation can be balanced, operation consistency can be improved, resource waste and low efficiency caused by improper task allocation can be avoided, risk level calibration of task units can be achieved, and the method is suitable for popularization and application. Potential problems are found in advance, effective countermeasures are taken, scheduling decision precision and flexibility are improved, and efficient and smooth execution of construction tasks is ensured.

The invention discloses a self-adaptive energy consumption optimization storage chip control method and device and a medium, and relates to the technical field of computer structures.The method comprises the steps that feasibility judgment is conducted on operation parameters in a candidate parameter set, reliability behavior prediction is conducted through a state evolution mapping relation, and an extension feasible set is formed; operation cost and maintenance cost comprehensive evaluation is carried out on the extended feasible set, multi-target optimization sorting is carried out, and a target operation parameter combination is output; performing prediction deviation analysis on the execution result packet, updating a prediction mapping relation and a historical parameter library, and outputting a self-adaptive control basis; according to the self-adaptive storage chip control method, the historical behavior data, the current junction temperature information and the current state distribution of the target storage partition are collected, so that the self-adaptive storage chip control which controls the model to continuously self-evolve along with the aging process of the chip and gives consideration to energy consumption optimization and service life reliability is realized.

The application discloses a unified probability propagation prediction method and system suitable for WUI fire, and the method comprises the following steps: constructing a unified state space and discretizing; constructing an external driving model of building layer fire propagation; constructing a building layer dynamic Bayesian network with each building as a node, parameterizing a conditional probability table of the building layer dynamic Bayesian network based on a physical driving mechanism; establishing a mapping relationship between an internal state of the building layer and an external observation category; realizing state correction and rolling prediction based on Bayesian updating; when external observation is obtained, the system state is updated based on Bayesian updating in combination with observation information, a posterior state distribution is obtained, the posterior state distribution is taken as an initial state of the next time step prediction, a rolling assimilation closed loop is formed, and finally a multi-state probability risk field of building scale evolution over time is output. The application can realize unified coupling of a landscape, a building and a flying fire, obtain state probability output over time, and can be assimilated and corrected.

The application discloses a content recommendation method and related products, which comprises the following steps: obtaining a current intention state distribution of a first object; predicting a predicted intention state distribution corresponding to the first object according to a historical intention state sequence of the first object and a current interaction environment feature; performing multi-dimensional confidence evaluation on the first object according to the current intention state distribution, the current interaction environment feature and the predicted intention state distribution to obtain a comprehensive confidence; obtaining a current intention state recommendation score and a future intention state recommendation score corresponding to a target candidate content; performing weighted fusion processing on the current intention state recommendation score and the future intention state recommendation score according to the comprehensive confidence to obtain a final recommendation score corresponding to the target candidate content; and using the final recommendation score to determine a ranking position of the target candidate content in a content recommendation set.

The invention discloses an aeration dynamic equilibrium digital control method and system coping with flow state distortion, and relates to the technical field of sewage treatment.The method comprises the following steps that a sewage treatment aerobic pool multi-point dissolved oxygen and flow state synchronous monitoring system is constructed, spatial distribution data of dissolved oxygen, water flow velocity and density in a pool are continuously collected, and the spatial distribution data are calculated; generating a flow state distribution map reflecting flow state distribution characteristics in the pool; and carrying out cooperative feature extraction on the collected water flow velocity gradient and dissolved oxygen difference based on the flow state distribution map, and identifying an abnormal region with flow state distortion. Through multi-point monitoring and dynamic weighting control, real-time identification of the flow state of the aerobic tank and accurate adjustment of the air volume are realized, the distribution balance of dissolved oxygen is maintained, and the oxygen transmission efficiency and the operation stability are improved; by introducing a fluid correction factor and a time sequence prediction model, a control closed loop with self-learning and energy-saving optimization capabilities is established, and the intelligence and economy of the aeration process are improved.

The present application relates to the technical field of remote scheduling, in particular to a kind of roadbed pavement construction equipment remote scheduling method and system, comprising the following steps: based on task unit division, identify device trajectory inflection point and operation state conversion mutation interval, analyze device running direction consistency and operation state distribution, assess linkage task scheduling efficiency, analyze construction area stress distribution trend, calibrate abnormal grade, identify corresponding number, output construction equipment intelligent collaborative scheduling instruction set.In the present application, by accurately extracting and analyzing the running track and operation state of the equipment, the abnormal interval and deviation area can be effectively identified, which helps to balance resource allocation, improve operation consistency, avoid resource waste and low efficiency caused by improper task allocation, and realize risk level calibration of task unit, so as to find potential problems in advance and take effective measures, improve scheduling decision accuracy and flexibility, and ensure efficient and smooth execution of construction tasks.

The invention provides a time series data synthesis method and device and a storage medium, relates to the technical field of data processing, and can improve the accuracy of a time series data synthesis result. The method comprises the steps of determining a plurality of patch sequences based on a preset window and input data; the input data comprises a plurality of data sequences and a category label of each data sequence in the plurality of data sequences, and the plurality of patch sequences are obtained by intercepting each data sequence through a preset window; based on each patch sequence and a class condition semi-Markov rule, state parameters of each data sequence under each class label are determined, and the state parameters comprise at least one of initial state distribution, a transition probability matrix and state duration time distribution; based on the state parameters and the input data, a synthetic data sequence of the input data is determined.

The invention discloses a convergence control method for a job event flow path. The method comprises the following steps: serializing each process node and forming potential energy distribution mapping; constructing a path force field of the reference path; mapping the multi-dimensional state parameters of the to-be-processed job event to a path force field, and generating a drift metric; calculating a potential energy change function, and dividing the to-be-processed event into at least a steady-state region, a metastable-state drift region and an unsteady-state distribution region; determining disturbance sources influencing the potential energy track of the to-be-processed event, and determining a coupling strength parameter and a disturbance projection direction vector of each disturbance source; superposing at least one compensation sub-force field on the basis of the path force field to generate a corrected potential energy field; calculating a potential energy gradient function and a minimum potential energy value of the event to be processed, and generating an event convergence degree function; and performing mapping comparison on the event convergence degree function and a preset threshold value set, and generating a control instruction set according to a comparison result.

This invention relates to the field of space debris environment modeling technology, specifically to a short-term evolution method for continuum disintegration debris based on numerical methods. The method includes: Step 1, establishing an initial distribution model of disintegration debris, including a disintegration debris velocity increment model and a numerical model of the initial distribution of disintegration debris; Step 2, sampling feature points of the continuum using a gradient-based sampling method, selecting the top N samples with the highest L2 norm scores of the gradient vectors as feature points of the continuum to be evolved; Step 3, processing the continuum equations using numerical solutions, considering the six orbital roots and four perturbation factors to perform short-term evolution of feature points; Step 4, reconstructing the final state distribution of disintegration debris using a Student's t-mix model. This invention fills the gap in previous research in this field, which only focused on ultra-short-term and long-term evolution, and has the beneficial technical effects of requiring fewer samples, having lower computational costs, shorter computation time, higher result accuracy, and stronger robustness.

The invention provides a micro-channel fluid simulation method and system, and the method comprises the steps: constructing a D2Q9 model through a lattice Boltzmann method according to a to-be-simulated micro-channel geometric structure, setting basic parameters of the D2Q9 model, and initializing distribution functions of a flow field, a temperature field and a phase field and a corresponding equilibrium state distribution function; respectively executing collision and propagation operations on distribution functions of a flow field, a temperature field and a phase field based on a single relaxation lattice Boltzmann method, and simulating movement among fluid particles, energy transmission and evolution of a two-phase interface; setting boundary conditions of the D2Q9 model, and calculating a macroscopic physical quantity representation simulation result of the fluid at each moment according to an evolution result of each distribution function after collision propagation; and when the D2Q9 model meets a preset simulation termination condition, outputting a D2Q9 model simulation result in a standard data format. According to the scheme, the accuracy of micro-channel internal flow simulation can be improved, and the simulation calculation efficiency is guaranteed.

The invention discloses a machine learning-based dynamic regulation and control method and system for the utilization ratio of a sintering return mine cold-pressing block blast furnace, and relates to the technical field of blast furnace ironmaking, and the method comprises the following steps: collecting and preprocessing blast furnace multi-source process parameters in real time, calculating the deviation degree between the selected parameters and a forward motion reference value, and generating parameter state distribution feature vectors; fusing the feature vector and original process data, inputting the fused feature vector and original process data into a pre-trained machine learning model, outputting an optimal cold pressing block charging proportion, and converting a proportion instruction into a control signal of a feeding system through a controller to realize proportion dynamic adjustment; the system comprises a data acquisition and preprocessing module, a parameter state feature generation module, an intelligent decision module and an instruction execution regulation and control module. According to the method, the parameters reflecting the operation state of the blast furnace are calculated and classified, the input quality and decision accuracy of the model are remarkably improved, real-time and self-adaptive accurate regulation and control of the cold pressing block proportion are achieved, and reasonable utilization of resources is promoted while smooth operation of the blast furnace is guaranteed.

The application discloses a kind of based on look-up table's memristor conductance regulation method and its application. Among them, the conductance regulation method includes the following steps: when the conductance of memristor needs to be regulated to target conductance interval, the pulse sequence information matched with target conductance interval is found from the look-up table made in advance, and the look-up table stores the corresponding relationship between at least two conductance states and corresponding pulses;According to pulse sequence information, corresponding pulses are applied between the upper and lower electrodes of memristor, and the regulation of the conductance of memristor is completed.The application statistically analyzes the conductance state distribution of the device, manufactures the look-up table, programs and regulates the conductance of memristor using the look-up table method, which can effectively reduce the number of read operations, and different conductance states correspond to known pulse design and pulse sequence design, without repeated verification whether to reach target conductance interval, the programming process is simple, and since all conductance states are known, there is no situation that cannot be written into the conductance interval during the programming process.

The tracking device (1) for tracking a target (2) comprises: a prediction module (100) configured to predict a state distribution of the target (2) at a certain time k; a correction module (120) configured to define an accuracy of the state distribution predicted by the prediction module (100) by a likelihood function at at least one observation point of the target (2) observed at the certain time k, and correct the likelihood function according to a number M of the observation points; and an update module (140) configured to update the state distribution at the certain time k based on the likelihood function corrected by the correction module (120).

The invention discloses an unmanned aerial vehicle communication multi-priority queue modeling and performance analysis method, and the method comprises the steps: building an extensible state model through a WiMarkov chain and a hypercube transfer unit, and achieving the low-delay access of different priorities through a threshold and a window in combination with a weighted COS statistics and fixed backoff mechanism; semi-analytical expressions of the success rate and the time delay are obtained through steady-state distribution and generation function approximation, and parameters are learned through optimization constraints. Simulation shows that the low-load success rate is larger than or equal to 99%, the high-priority delay is still kept at the millisecond magnitude during high load, and the method is suitable for an unmanned aerial vehicle network communication system with the strict real-time requirement.

An interface state testing method and a reliability testing method belong to the technical field of power device testing, and the interface state testing method comprises the following steps: applying a plurality of cluster pulse voltages to a grid electrode of a device to be tested according to a time sequence, and obtaining a charge pump current corresponding to each cluster pulse; obtaining a charge pump curve of the to-be-tested device according to the charge pump current corresponding to each cluster of pulses; and according to the charge pump curve of the to-be-tested device, calculating the average value of the interface state distribution of the to-be-tested device. According to the invention, bias voltage does not need to be applied to a drain-source loop, and the interface trap information of a single vertical MOSFET device can be accurately extracted by controlling the gate pulse and measuring the drain current.

This invention discloses an adaptive path planning method for intelligent unmanned vehicles in dynamic environments, relating to the field of path planning technology. The method includes: determining whether the comprehensive confidence index of the current potential state representation is lower than an adaptive threshold based on state distribution data; if lower than the adaptive threshold, dynamically generating multiple parallel imagined branches in the potential space according to the degree of confidence deficiency, and continuously predicting the future potential state sequence corresponding to each imagined branch; based on the future potential state sequence corresponding to each imagined branch, sampling candidate control sequences in parallel on multiple imagined branches, and selecting the optimal control command based on the safety performance of each candidate control sequence under the evolutionary path formed by different imagined branches; mapping the optimal control command to a physical space trajectory and driving the unmanned vehicle; thus improving the robustness and interaction safety of path planning for unmanned vehicles in complex and uncertain environments.

The invention discloses a multi-task deep reinforcement learning system based on a mixed world model. The system comprises a multi-task environment model, an agent strategy and a mixed world model. Wherein the framework of the mixed world model is designed by mixing a plurality of experts Transformers and connecting the experts Transformers in series to share the Transformers, and each expert Transformer is activated through task condition routing selection and is used for selecting an intelligent agent according to an intelligent agent action input by an intelligent agent strategy and high-dimensional environment data input by a multi-task environment model, so that the multi-task environment model is constructed according to the intelligent agent action input by the intelligent agent strategy and the high-dimensional environment data input by the multi-task environment model. State distribution, rewards, termination signals and task identifiers of the next time step are predicted. After training the mixed world model in an end-to-end mode, generating imaginary trajectory data in an autoregression mode; and training the intelligent agent strategy according to the imaginary trajectory data. According to the method, the generalization ability and the sample efficiency of the intelligent agent among heterogeneous tasks can be enhanced, and cross-task high-fidelity state prediction and efficient strategy learning are realized.

Embodiments of the present application disclose a state analysis method and device, computer equipment and a storage medium. The method comprises: acquiring video data collected by a wearable device during wearing; performing feature extraction on the video data to obtain scene features, object features, limb features and visual angle motion features; generating target features according to the scene features, object features, limb features and visual angle motion features; inputting the target features into a trained state distribution identification model to output a current state feature corresponding to the video data and a corresponding standard state, wherein the trained state distribution identification model is trained based on label states of sample video data corresponding to an object, and sample scene features, sample object features, sample limb features and sample visual angle motion features in the sample video data; determining a state difference between the current state feature and a standard state feature corresponding to the standard state, and determining a current state of the object according to the state difference.

The invention relates to the technical field of three-dimensional digital modeling, and discloses a three-dimensional Gaussian particle modeling method, electronic equipment and a readable storage medium. The modeling method comprises the steps of multi-modal data space embedding and fractional order space prior construction, state-guided three-dimensional Gaussian particle initialization and attribute expansion, state-aware projection rendering mechanism design and forward propagation modeling, multi-task particle attribute optimization mechanism under joint supervision and three-dimensional state field inference and output. Fractional order space prior is introduced, so that multi-modal information can be embedded into state particles in a non-local mode, and deep coupling modeling of a structure and an operation state is achieved. A state-aware particle rendering mechanism is introduced, synchronous expression of three attributes of space, color and state is achieved, and the expression ability of three-dimensional state distribution is improved; and a structure-state joint regular optimization strategy is introduced, so that the spatial discontinuity of state expression is effectively inhibited, and the physical consistency and interpretability of a three-dimensional state field are ensured.

The application provides a quantum flow simulation method and device based on LBM, a medium and equipment, which can solve the problem of large amount of calculation of classical LBM in related technologies, improve the Reynolds number simulated by the simulation method, and widen the practical application range of the simulation method. The method comprises the following steps: initializing flow field information; using four quantum bits to encode the flow field information and the distribution function based on a two-dimensional nine-velocity discrete model; obtaining a first-order equilibrium state distribution function in each discrete velocity direction based on the encoded flow field information; performing a collision operation; performing a migration operation; updating the flow field information based on the migration operation; judging whether the updated flow field information meets a convergence condition; if not, performing boundary condition processing on the updated flow field information, and returning to the step of encoding the flow field information and the distribution function based on the processed flow field information until the updated flow field information meets the convergence condition, and outputting the updated flow field information.

The application relates to a Markov chain and regression analysis-based application system availability prediction method, and belongs to the field of fault prediction. The application determines an index set influencing the availability probability of an application system, defines the state space of each index in the index set, determines the relationship between the state space of each index and the availability probability of the application system by using a regression analysis method, constructs a model for predicting the state distribution of each index by using a Markov chain model, determines the initial state of each index, defines the time range of the availability probability prediction of the application system, predicts the state distribution of each index of the application system, and predicts the availability probability of the application system after a specified time according to the state distribution of each index and the relationship between the state space of each index and the availability probability of the application system. The application effectively improves the availability of the application system, prolongs the stability time length of the application system, and improves the user satisfaction.

The invention discloses a nonlinear filtering method and a nonlinear filtering system based on mode matching, which are used for effectively solving the problem of state estimation of additive and asymmetric skew noise in a system model. The core of the method is to construct a mode matching (GeMM) transformation module, through synchronous processing and accurate propagation of three statistics of a mean value, a covariance and a mode of state distribution, and through matching by using closed skew normal (CSN) distribution, skew characteristics are accurately described. The specific implementation comprises the steps of system modeling and filter initialization; in the prediction stage, a prior mean value and a prior covariance are calculated by adopting a traditional moment propagation method, and a prior mode is directly and accurately calculated according to a mode propagation theorem; in the updating stage, the prior statistic, the observation noise statistic and the actual observation value are input into GeMM for transformation, the transformation finally obtains the mean value, the covariance and the mode of posterior condition distribution by constructing CSN distribution of the statistic matched with the joint variable, and state estimation is completed. The method provided by the invention can be used as an enhanced module seamless embedded extended Kalman filter (EKF) or unscented Kalman filter (UKF) framework, the estimation precision in a skew noise environment is remarkably improved, excellent calculation efficiency and robustness are kept, and the method has a good engineering application prospect.

The invention discloses a video adaptive transmission method based on a recurrent neural network, and the method comprises the following steps: S1, collecting data, and generating a network state sequence and a buffer region sequence; s2, detecting playing session switching, calculating a migration weight based on a historical abstract and a current network state sequence, and initializing SRNN hidden state distribution parameters; s3, inputting the network state sequence into the SRNN model, and outputting network state prediction distribution parameters; s4, calculating an underflow risk index and write-back intensity, and writing back and updating the hidden state distribution parameters; s5, calculating an evaluation value according to the network state distribution parameters and the buffer sequence, and determining a target code rate; s6, performing hierarchical budget evaluation on the parameter configuration, and screening candidate parameter configuration; and S7, failure configuration is judged according to the performance sequence, screening is constrained, optimal parameter configuration is output, and transmission is executed circularly. According to the method, network state prediction and buffer risk cooperative control are realized, and the video playing stability and the adaptive decision reliability are improved.

This invention provides a method and system for controlling the production of silicone sealant based on digital twins. The method includes: simulating the flow field evolution in a digital twin stirring model using impeller motion state data as motion boundary conditions, combined with static boundary conditions and initial flow field conditions, to obtain the flow state distribution relationship of the sealant within a physical stirring device; constructing a state evolution model of the sealant during the mixing process based on the flow state distribution relationship; predicting the stirring intervention strategy indicated when the sealant reaches the target mixing state based on the state evolution model; controlling the physical stirring device to perform operations based on the stirring intervention strategy; updating the motion boundary conditions based on the operating parameters of the physical stirring device after the operations; simulating the flow field evolution and state evolution prediction based on the updated motion boundary conditions; iteratively optimizing the stirring intervention strategy until the target mixing state is met, at which point the physical stirring device is stopped. This invention improves the product quality of silicone sealant during the production process.