37results about How to "Quick response to changes" patented technology

The invention discloses a mobile game scene interaction intelligent optimization method based on a generative adversarial network, and aims to improve the game experience of players. Comprising the following steps: S1, collecting behavior data and interaction information of a player in a mobile game, and generating a preliminary behavior feature model; s2, constructing a generative adversarial network model; s3, generating diversified game scene interaction examples according to the trained generator; s4, evaluating and optimizing the generated scene interaction example by using a discriminator; s5, real-time feedback of the player is received, and the behavior feature model of the player is updated; s6, dynamically adjusting a generation strategy based on the updated behavior feature model, and optimizing a scene interaction effect; s7, environment data are integrated, scene interaction is optimized in real time, and the immersion of players is enhanced; and S8, providing user-defined scene interaction options for the players, and allowing the players to adjust according to personal preferences. The interactive experience and game fun of players can be effectively improved.

The invention belongs to the field of medical information systems, and provides a medical institution full-process doctor-seeing path planning method and system based on artificial intelligence, and the method comprises the steps: modeling each doctor-seeing link in a medical institution into a directed graph; the node features and the edge weights of the directed graph serve as input, the node features are updated through a graph neural network, and the weight of each node is dynamically generated; a queuing theory model is adopted for modeling according to the queuing condition of each node, and the average waiting time of each node is calculated; designing a multi-objective optimization function, and taking balance of the shortest path and the minimum waiting time as an optimization objective to generate an optimal doctor-seeing path scheme; the resource state of each node is dynamically monitored based on a real-time feedback mechanism of the Internet of Things equipment and a hospital information system, and when it is monitored that the resource state or waiting time of the node is changed, the graph neural network and the queuing theory model are recalled to update node features and priority weights.

The invention relates to a multi-language intelligent question and answer method and device, equipment and a medium in the technical field of e-commerce, and the method comprises the steps: obtaining a question text provided by a user, and determining a question language of the question text; when the question language is consistent with a knowledge language corresponding to the question and answer knowledge base, determining a pre-stored question matched with the question text in the question and answer knowledge base, and taking an answer text corresponding to the pre-stored question as a target answer text; when the question language is inconsistent with the knowledge language, translating the question text into a target question text corresponding to the knowledge language, determining a pre-stored question matched with the target question text in a question and answer knowledge base, and translating an answer text corresponding to the pre-stored question into a target answer text corresponding to the knowledge language; and sending the target answer text to the user. Only one question and answer knowledge base of a specific language needs to be maintained, and the management and maintenance workload of the question and answer knowledge base is remarkably reduced.

The invention belongs to the technical field of vehicles, and particularly relates to a steering angle module chassis and a vehicle. The steering angle module chassis comprises a frame platform, a steering angle module and a hydraulic interconnection suspension system. The steering angle modules are arranged at the four corners of the vehicle frame platform and correspond to the left front steering angle module, the right front steering angle module, the left rear steering angle module and the right rear steering angle module respectively. Each steering angle module comprises a vibration reduction sliding column assembly. The hydraulic interconnection suspension system is arranged on the vehicle frame platform and connected to the vibration reduction sliding column assemblies in the steering angle modules so as to be used for adjusting and controlling the posture of the vehicle body. According to the hydraulic interconnection suspension system, the interconnection relation and internal hydraulic pressure of the vibration reduction sliding column assemblies in the four steering angle modules can be changed, four-wheel movement interconnection is achieved, the change response to the road surface is better and more rapid, then the inclination angle of a vehicle body is controlled, and the purpose of regulating and controlling the posture of the vehicle body is achieved.

The invention discloses a data security compliance evaluation and identification method and system, and relates to the technical field of data security compliance evaluation and identification, and the method comprises the steps: collecting regulation text data through an API interface, and carrying out the preprocessing of the regulation text data through a natural language processing library NLTK, and obtaining formatted regulation text data; mapping each word in the formatted law and regulation text data into a fixed-length vector by adopting a pre-calculation word embedding method, and generating a digital vector; marking important terms in the digital vector by adopting a named entity recognition NER method to obtain a key concept text; establishing an evaluation model based on a machine learning algorithm, inputting the digital vector into the evaluation model, calculating and outputting an evaluation value by adopting a weighted average method, setting an evaluation threshold value, comparing the evaluation threshold value with the evaluation value, judging whether the digital vector is compliant or not, and obtaining an evaluation result; and constructing a recognition model based on an anomaly detection algorithm, and inputting the key concept text into the recognition model.

The invention relates to the technical field of motor control and parameter identification, and discloses a permanent magnet synchronous motor parameter identification method based on a self-adaptive forgetting factor, which comprises the following steps of: acquiring an input signal and an output signal of a permanent magnet synchronous motor in real time, and constructing a regression vector by using the acquired input signal and output signal; calculating the information entropy of the current moment according to the constructed regression vector; performing normalization processing on the information entropy, and calculating a forgetting factor at the current moment according to the normalized information entropy; and substituting the forgetting factor into a parameter updating formula of a recursive least square method for parameter identification. According to the method disclosed by the invention, the adaptive capacity to dynamic change of motor parameters can be improved, the interference of noise on identification precision is effectively suppressed, the credibility of an identification result is improved, and the robustness and long-term stability of a control system are ensured.

The invention discloses an energy regulation and control method and system of a virtual power plant based on a prediction model, and relates to the field of virtual power plant operation control, and the method comprises the following steps: collecting energy data in the virtual power plant; integrating the wind power data and the load data, and performing feature extraction on the integrated data; based on the predicted power generation data of the distributed energy and the predicted demand data of the load side resources, constructing a preliminary energy dynamic balance prediction model; performing day-ahead load demand and distributed energy prediction by using the preliminary energy dynamic balance prediction model; according to the preliminary scheduling strategy, using a niche grey wolf optimization algorithm to adjust the operation state of each distributed energy and load side resource in real time; dynamically adjusting and optimizing the energy dynamic balance prediction model; and carrying out load demand and distributed energy prediction again by using the optimized energy dynamic balance prediction model. According to the method, the change of the power system can be responded in advance, and more efficient energy scheduling is realized.

The invention relates to the technical field of intelligent eye medicine boxes, and discloses an intelligent eye medicine box and a control method thereof.The eye medicine box is divided into a plurality of compartments for storing different types of medicine, and a sensor is arranged in each compartment to detect the weight of the medicine in real time; and accurately calculating the residual amount of the medicine by combining the initial weight data of the medicine, the dosage of the medicine currently used by the user, the last time of using the medicine by the user, the weight of the medicine in the last time and the total number of times of using the medicine by the user. Furthermore, the future drug use dosage and use rate are predicted through a machine learning model, the remaining amount is optimized based on the prediction results and the real-time weight data, the optimized remaining amount is obtained, and the optimal remaining amount is calculated according to the optimized remaining amount and the future drug use rate. Specifically, firstly, the accuracy of medicine management is remarkably improved through a smoke and flame identification method, it is ensured that a patient can use medicine on time according to the amount, and the poor treatment effect caused by missed taking or wrong taking is avoided;

The invention discloses an on-chain enterprise risk identification method and system based on a credit chain large model. The system comprises a data preprocessing module, a model matching module, a weight parameter loading module, a data operation reasoning module, a risk prediction output module and the like. Large model training and optimization are carried out by integrating industry, region, type and other multi-dimensional data sets, a credit chain vertical type general model and a special model are formed, and the method can adapt to credit and risk prediction evaluation, monitoring and early warning of a single enterprise or an integrated enterprise in different supply chain and industry chain scenes. A more comprehensive, accurate and real-time solution is provided for enterprise management and industry supervision, and remarkable beneficial effects and advantages are achieved.

The invention provides a railway green performance data knowledge extraction method. The method comprises the steps of data collection and preprocessing, knowledge graph construction and data standardized storage. The method comprises the following steps: collecting and classifying key data of a full life cycle of a railway; constructing an ontology by a top-down method, extracting knowledge, and storing the knowledge in a graph database; storing the structured data based on the metadata, sorting and classifying the unstructured data, binding the unstructured data with the knowledge graph, and storing the unstructured data; according to the method, the environment-friendly degree of railway engineering can be improved, information management is optimized, the data driving force of decision support is increased, data sharing and communication are promoted, the sustainability of railway projects is improved, compliance of environmental protection rules and policies is achieved, and the timeliness and dynamic updating capacity of data are improved; not only are knowledge extraction efficiency and accuracy improved, but also monitoring and improvement of key performance indexes of the railway engineering are more scientific and systematic, and energy conservation, consumption reduction and ecological environmental protection of the railway engineering are further promoted.

The invention discloses an auxiliary driving control method and device through a construction road, equipment and a storage medium, and relates to the technical field of intelligent driving, and the method comprises the steps: obtaining cloud construction information and perceived road condition information; performing credibility judgment on the cloud construction information based on the sensed road condition information; when the judgment result is credible, performing collision prediction on the vehicle based on the sensed road condition information and the cloud construction information; and when the collision prediction result is that collision possibly occurs, auxiliary control is performed on the vehicle, and the change of a construction road can be quickly responded, so that accidents are reduced and the driving experience is improved.

The invention discloses a knowledge-embedded fusion-driven power grid dispatching adaptive cruise decision-making method and system, and the method comprises the steps: building a power grid dispatching adaptive cruise decision-making problem mathematical model, obtaining the operation state prediction data of a future time period, and carrying out the scheduling plan making and dynamic optimal trend adjustment; establishing a power grid dispatching adaptive cruise decision offline training environment based on deep reinforcement learning, and constructing a state space, an action space and a reward function; and constructing a near-end strategy optimization algorithm of knowledge embedded fusion for guiding optimal learning of the scheduling intelligent agent, and obtaining an optimized scheduling plan generated by the scheduling cruise intelligent agent. Through the deep reinforcement learning technology based on knowledge embedded fusion driving, the large power grid dispatching adaptive cruise decision-making method is provided, the actual change of the power grid can be quickly responded, the power grid dispatching precision is improved, higher-level automation and intelligence are achieved, and the power grid dispatching efficiency is improved. And the method has relatively high practical value for promoting the intelligent level improvement of power grid dispatching operation.

The invention provides a penetration test method and system, and the method comprises the steps: obtaining an information updating event, judging the type of the information updating event: when the information updating event is an inherent information updating event, selecting and executing a corresponding inherent test task according to the updating content of the inherent information updating event, after the inherent test task is completed, updating the corresponding asset business information, and issuing a business information updating event; when the information updating event is a business information updating event, selecting and executing a corresponding business test task according to the updating content of the business information updating event, and entering a post penetration process after the business test task is completed; and after entering the post-penetration process, performing a post-penetration test on the corresponding electronic equipment. According to the test method and the test system for the penetration test, the efficiency and the safety of the penetration test can be improved.

The invention belongs to the field of microelectronic metering testing, and particularly discloses an integrated circuit standard sample wafer and a testing method thereof, and the standard sample wafer comprises a substrate, and a system-on-chip SoC, an analog-to-digital converter, a digital-to-analog converter, a reference source and a data selector which are mounted on the top surface of the substrate; the SoC is used for controlling the output, acquisition and feedback control of a standard signal; the analog-to-digital converter is used for realizing acquisition and closed-loop feedback of standard signals; the digital-to-analog converter is used for realizing output and closed-loop feedback of a standard signal; the reference source is used for providing a reference voltage of the analog-to-digital converter and a reference voltage of the digital-to-analog converter; and the data selector is used for realizing channel switching of the data test system so as to realize multi-channel test of a single standard sample. According to the invention, the test work efficiency of the integrated circuit test system can be improved.

The present invention relates to the technical field of drought remote sensing monitoring, and specifically discloses a drought remote sensing monitoring method and system integrating multi-source remote sensing data, the method comprising: dividing the area to be measured into a plurality of position units; obtaining the temperature, altitude and air pressure data of each position unit in the area to be measured, and calculating the saturated water vapor pressure difference of each position unit; simultaneously obtaining the soil moisture and sunlight-induced chlorophyll fluorescence data of each position unit in the area to be measured; performing data normalization; selecting the driest position unit as the reference position unit, comprehensively calculating the saturated water vapor pressure difference, soil moisture and sunlight-induced chlorophyll fluorescence dimensions, calculating the distance between other position units and the reference position unit, and obtaining the drought index of each position unit; and using the drought index to realize drought monitoring of different positions in the area to be measured. The present invention can more accurately quantify the degree of drought, comprehensively extract the comprehensive information of drought events, and has good reliability and accuracy.

The invention relates to the field of non-contact physical sign perception, in particular to a human body vital sign perception method based on VMD + ACMD joint decomposition, which comprises the following steps of data acquisition, distance angle heat map generation, direct current bias removal, expansion DACM phase extraction, signal decomposition, signal reconstruction, feature extraction, time-frequency analysis and the like. According to the invention, the VMD and the ACMD are combined for use, and the frequency resolution capability of the VMD and the adaptive decomposition capability of the ACMD are fully utilized. Through the combined decomposition, the analysis capability of the micro-motion signal is enhanced, the sign extraction precision is remarkably improved, compared with a traditional single decomposition method, the scheme improves the separation precision of the sign signal, the modal aliasing problem is reduced, the higher signal quality is ensured, and by enhancing the main frequency component of the sign signal, the sign extraction precision is improved. Interference of background noise is effectively suppressed, especially in a complex or high-noise environment, high signal-to-noise ratio of signals can be ensured, and accurate extraction of sign signals such as heart rate and respiratory rate can be improved.

The invention discloses a Hash and feature selection-based software and hardware equipment fingerprint fine-grained generation method, and belongs to the technical field of power network software and hardware equipment fingerprint biomarker generation. Based on the information entropy and the self-correlation coefficient, the features used for generating the device fingerprints are screened out from the numerous features, representative and stable features are extracted from the device fingerprints, and selected feature parameters are converted into hash codes used for the device fingerprint generation and matching step; different devices are identified and distinguished hierarchically and accurately through a fine-grained device fingerprint generation step based on feature selection, a hash code and a Bayesian network.

The invention relates to the technical field of industry management, in particular to a carbon-based conductive new material industry brain management cloud platform. The method comprises the following steps: a data acquisition unit is connected to equipment for producing the carbon-based conductive material to collect production parameters, and uploads the parameters to a cloud platform; the data storage management unit processes and stores different types of data through distributed storage; the data analysis unit utilizes a support vector machine model to distinguish the quality of the produced carbon-based conductive material, estimates the yield based on a ridge regression model, applies a particle swarm optimization algorithm to improve the production efficiency, and introduces an influence function of porosity in the distinguishing process of the support vector machine model. According to the design, the support vector machine model is introduced, the kernel function is optimized according to the characteristics of the carbon-based conductive material, particularly, the influence of the porosity and the surface gas adsorption quantity on the conductivity of the material is considered, and the platform can more accurately distinguish the quality of the production material.

The utility model discloses a kind of optical fiber boxes convenient to disassemble, through the unique cooperation mode of J-shaped hook and fastening hanging hole, the quick, non-destructive disassembly of drawer inner box is realized.When installing, by lifting upward and pushing inward drawer inner box operation, the resistance of J-shaped hook is successfully passed through fastening hanging hole and finally firmly hooked backboard, the process does not need any tool and action is smooth.This structure ensures the stability of connection after installation, effectively prevents drawer from accidentally falling out due to vibration or misoperation in drawing use.When disassembling, drawer inner box can be removed by simply lifting and pulling, which greatly facilitates the maintenance and replacement of internal optical fiber devices.This intuitive and reliable operation mode significantly improves the operation efficiency, reduces the operation difficulty and the wear of components, and prolongs the service life of equipment.

The invention discloses a transient frequency support control method considering the low voltage penetration characteristic of a converter, and relates to the field of electrical engineering, and the method comprises the steps: collecting the voltage data and frequency data of a power grid through a real-time monitoring sensor; based on the acquired power grid voltage data, carrying out converter voltage low-pass-through characteristic detection to obtain a voltage low-pass-through state; according to the voltage low penetration state, starting an adaptive response strategy, and adjusting a power instruction; based on the power grid frequency data, starting a virtual synchronous machine control algorithm according to the adjusted power instruction, and adjusting power output; a reinforcement learning algorithm is used to carry out cooperative adjustment on the voltage low-penetration adaptive response strategy and the virtual synchronous machine control algorithm, and the state of the power grid is recovered; according to the invention, through cooperative adjustment of the self-adaptive response strategy of the voltage low penetration state and the virtual synchronous machine control algorithm, the effect of the converter in power grid frequency adjustment is effectively improved, and the stability and reliability of the power grid frequency are enhanced.

This invention discloses a power supply control system for charging piles based on photovoltaic energy storage, relating to the field of charging pile power supply technology. It solves the technical problem in existing technologies where disordered charging behavior (such as random charging and large fluctuations in charging duration) leads to irregular fluctuations in the load on the power supply network. This invention achieves intelligent, personalized, and efficient operation of charging pile power supply through refined characteristic identification, accurate status monitoring, and dynamic power supply decision-making. In terms of energy utilization, the system significantly improves photovoltaic energy utilization efficiency by accurately matching photovoltaic power supply with charging demand, reducing dependence on grid power and minimizing energy waste. Regarding grid security, by accurately identifying grid damage risks and load interference states, it achieves early warning and response to grid risks, reducing the risk of grid overload operation, extending the service life of power supply equipment, and improving the safety and stability of grid operation.

The invention relates to the field of isolation switch driving control, in particular to a demodulation device for isolation switch driving control and a driving method of the demodulation device. Waveform shaping is carried out on an input modulation signal, the modulation signal is converted into a high-frequency square wave signal, high-frequency components are filtered out through a low-pass filter, and a stable digital signal is obtained; the method comprises the following steps: carrying out delay processing on a modulation signal containing logic jump to ensure that the delay width can cover the low level width of the signal, controlling a clock of a delay timer through a control logic module according to the low level part of the modulation signal to avoid generating a rising edge signal, and outputting a reset pulse signal and a driving low signal when a specific condition is met. When other conditions are detected, the register is triggered to sample a high-level signal and output a driving high signal, interference can be effectively filtered, the amplitude and the phase of the signal are kept stable, and low delay is achieved in the signal conversion process so as to ensure the accuracy of signal transmission.

The invention relates to the technical field of resonance suppression, in particular to a resonance suppression method, system and equipment for distributed energy storage network construction type control. By obtaining the stable integral gain and the stable differential gain, the response of the PID controller in the control process can be ensured to be more agile, the dynamic performance of the system can be optimized, the appropriate stable integral gain is beneficial to eliminating steady-state errors, and by obtaining the response characteristic curve in real time, the gain can be dynamically adjusted, the change can be quickly responded, and the resonance risk can be reduced. The amplitude-frequency characteristic curve and the phase-frequency characteristic curve provide visual evaluation for the stability and the response speed of the energy storage system, so that the PID control performance can be effectively improved, the risk of a resonance phenomenon is reduced, the stability of the energy storage system can be evaluated in real time by judging the stability margin, potential problems can be found in time, resonance is avoided, and the reliability of the system is improved. According to the judgment result of the stability margin, a lag compensator or a phase compensator is selected for adjustment, and the resonance problem in the energy storage system is solved.

The present invention discloses a server-side adaptive SVC ABR method based on MPQUIC and pre-fetching, wherein the client sends multiple video requests concurrently, the server extracts application layer information, and calculates the size of the client video buffer and the available bandwidth; then selects the corresponding video layer transmission strategy and makes a video stream selection decision; after the server sends the video stream to the client, when the video playback of the client freezes, the server dynamically adjusts the cutoff time of subsequent video clips until the entire video is played. The present invention can quickly respond to changes in the network environment and improve the bandwidth resource utilization of each sub-path. Compared with the existing ABR solution, the present invention can improve the overall video QoE by 40%, the average bit rate by 28%, and the average quality switching times are reduced by 15% compared with the existing ABR solution.

The utility model discloses a pair of outdoor trousers with a trouser leg positioning structure, which comprises a trouser body which is arranged below the trouser body, is bent inwards and is sewn and fixed to form a restraining channel; a connecting rope is installed in the restraining channel, and a connecting rope layout part extends out of the restraining channel to the outer side face of the trouser body to form a positioning part; the snap fastener is matched with the positioning part; by arranging the restraining channel and the connecting rope, the trouser legs can be effectively fixed through the connection of the positioning parts and the shoelaces, the trouser legs are prevented from shaking left and right in the walking or riding process, and discomfort and friction are reduced. Due to the fact that the positioning parts are matched with the snap fasteners and are tightly connected with the shoelaces, the trouser legs are prevented from being clamped by chains or riding equipment, and riding safety is improved. During riding, the trouser legs are fixedly connected with the shoelaces, so that the trouser legs are prevented from being blown up by wind or lifted up due to excessive action, the ankle parts are effectively protected, and the trouser legs are prevented from being rolled up.

The invention relates to an information processing method and device for content filtering and information classification, equipment and a medium, and the method comprises the following steps: calling a preset multi-dimensional classification model which automatically adjusts the classification weight according to the features of input data, and carrying out the multi-level feature extraction based on a GNN (Graph Neural Network) and a CNN (Convolutional Neural Network); and mapping the classification path to a processing tool Apache Kafka by using an automatic mapping mechanism, calling a signal processing module, a cross-domain ontology mapping module and a data flow analysis module, and deeply processing information. Through an automatic mapping mechanism, the system can dynamically map filtered information classification paths to a proper processing tool set, signal processing, cross-domain ontology mapping and data flow analysis modules are called, and the multi-dimensional depth information analysis capability ensures that the system not only can filter information, but also can analyze data flow. In addition, hidden association and deep semantic relationships among the information can be mined, and deeper insight is given to the information processing process.

The invention discloses an IoT (Internet of Things)-based low-voltage power distribution network loop closing monitoring system, which belongs to the technical field of power distribution network monitoring, and can ensure that the power grid states of two transformer areas are compatible before loop closing operation by monitoring key electric parameters such as voltage difference and phase angle difference in real time. Therefore, loop closing failure or power grid accidents caused by mismatching of the electric parameters are avoided. By automatically calculating the loop closing impact current, the steady-state current, the phase angle difference and the distribution transformer load rate difference, the system can provide a scientific judgment basis for operators, so that the loop closing operation is more accurate and reliable. Operators can know the state of the power grid in real time through the remote on-line monitoring system, so that the working efficiency is improved, and the safety risk of the operators is reduced at the same time. Through digital support, the system can quickly respond to the change of the power grid, timely adjust the loop closing operation, and improve the operation efficiency and power supply reliability of the power grid. The power failure time caused by loop closing operation is reduced through real-time monitoring and rapid judgment, and the power utilization experience of a user is improved.

The invention discloses a label management method for information system modeling. The method specifically comprises the following steps: label storage: storing a product and a label thereof into a namespace corresponding to a label source; label reading: inputting a target label identifier, traversing each namespace according to the priority from low to low, reading corresponding label data after finding the target label identifier, and stopping traversing; one product has at least one label, the signature identifier is used for identifying the corresponding signature, and the label data is used for describing the behavior of the product. Through priority management and a namespace mechanism, different products are allowed to dynamically manage and adjust marks in different scenes, so that system functions are more flexible, and complex and changeable service requirements are supported.

The invention discloses an air conditioner. The air conditioner comprises a load; the input end of the power circuit is connected with a single-phase power grid, and the output end of the power circuit is connected to the load; the power grid parameter detection module comprises an orthogonal signal generator which is used for converting the alternating-current input power grid voltage into orthogonal voltage signals V alpha and V beta under an alpha-beta shaft system; the Park conversion module is used for converting the orthogonal voltage signal into voltage components Vd and Vq under a synchronous rotating coordinate system; the per-unit phase locking module is configured to perform filtering on the voltage component Vd to obtain a power grid voltage amplitude Vs; performing per-unit processing on the voltage component Vq to obtain a per-unit value, and performing filtering and PI adjustment on the per-unit value in sequence to obtain a power grid voltage frequency omega; and performing integral adjustment on the frequency omega to obtain a power grid voltage phase theta. The power grid parameter detection module is used for power grid parameter information detection, and has the advantages of being high in anti-interference capacity and high in reliability.

The invention relates to the technical field of peanut harvesters, in particular to an automatic leveling system of a hill and mountain peanut harvester, which comprises a data acquisition module, a data preprocessing module, a self-adaptive gradient prediction and adjustment module, a dynamic control adjustment module and a power driving module, the data acquisition module acquires working state data of the peanut harvester in real time; the data preprocessing module is used for preprocessing the collected working state data; the self-adaptive gradient prediction and adjustment module dynamically adjusts the angle of a working platform of the peanut harvester; the dynamic control adjustment module performs real-time adjustment and dynamic correction and outputs a control instruction; the power driving module drives the electric hydraulic unit or the mechanical arm to conduct automatic adjustment. According to the invention, the stability and accuracy of the peanut harvester under various complex terrains are ensured, the risk of operation interruption caused by terrain changes is reduced, and meanwhile, the efficiency of harvesting operation is effectively improved.