37 results about "Decision cycle" patented technology

A video conference system based on Soft-MCU includes a server, a video conference terminal BHU-ICT, a computer terminals and video communication system procedures, wherein, the video conference terminal BHU-ICT is composed by a digital video camera, a microphone, and an audio and video encoder. The digital video camera and the microphone are respectively connected to the audio and video encoder, the encoder, the server and the computer terminal are connected on the network, and the procedures are respectively running in the server and the computer terminal. The digital video camera and the microphone acquire the audio and video information of the meeting place, and use the encoder to execute hardware coding. A media control procedure runs on the computer terminal, and an MCU service procedure run on the server terminal, uses the Ethernet to exchange the audio and video information with the encoder, for performing audio and video communications. All of the texts, requests, commands and other information formats in the system control signaling system conform to the XML standard. The invention constructs a remote network video communications platform with high-quality, high stability and high security, which can be widely used in party and government organs and enterprises and institutions. The invention has advantages of reduced system cost, reduced travel costs, shortened decision cycles and enhanced work efficiency.

A system and method have been provided for hierarchically arbitrating in a broadband information switching network. The method promotes the fair and efficient distribution of information packets across the switch fabric that ultimately permits the switch to maximally match information packets to switch output addresses, at faster rates and higher throughput. The method comprises: accepting variably sized information packets at a plurality of switch inputs, that address a plurality of switch outputs; at each switch input, queuing the information packets into a plurality of queues; parsing the information packets into units of one cell; simultaneously arbitrating between each switch output and a plurality of available switch inputs, where an available input is defined to have at least one queue with an information packet addressing that particular switch output; selecting a queue; locking the link between switch inputs and switch outputs; and, transferring information packets across the links in units of one cell per master decision cycle.

The invention discloses a multi-sensor cooperative tracking joint optimization decision method and aims at providing the optimization decision method with high calculating efficiency and an accurate optimization decision parameter. The method is characterized by predicting the target position information and the fusion measurement information of a sensor in a decision period by a multi-sensor cooperative tracking system, and obtaining the fusion measurement state of a decision cycle moment and a fusion measurement covariance; using multi-source data fusion algorithm to calculate the fusion tracking error of multi-sensor cooperative tracking to a target; according to local aircraft navigation information and radar detection information to the target, calculating the azimuth and the pitch angle of a local radiation source relative to the other aircraft and the tracking and positioning error of the local radiation source detected by the other aircraft; taking the fusion tracking error ofmultiple sensors to the target as a constraint condition and taking a target matching relationship and the like as an optimization decision parameter; establishing a multi-stage and multi-constrainedcooperative tracking decision model; and solving a multi-stage sensor optimization decision and the optimization decision parameter.

The invention relates to a method for use in a telecommunication system to generate an acceptance decision during a decision cycle that comprises whether to accept an incoming packet of a data stream or not in a buffer (BUF). The incoming packet (P) has a drop priority (DP) being one of a plurality of arranged drop priorities. The method comprises determining for a drop priority parameter (DPP) an actual drop priority parameter value (DPPact) out of the plurality of drop priorities; and comparing the drop priority parameter (DPP) with the drop priority (DP) of the packet (P) in order to generate the acceptance decision. The method further comprises that the step of determining the actual drop priority parameter value (DPPact) comprises determining a previous drop priority parameter value (DPPprev) by means of any one of an initial value (DPPinit) and a previous actual drop priority parameter value (DPPact′) having been determined during one of a previous decision cycle; and updating a queue occupancy variable (Q) as a function of a previous re-scaled queue occupancy value (RE-Qact′) having been determined during one of a previous decision cycle and providing thereby an intermediate queue occupancy value (Qint); and re-calculating the intermediate queue occupancy value (Qint) of the queue occupancy variable (Q) in function of instantaneous queue occupancy information (INF-Qins) of the buffer (BUF) and providing thereby an actual queue occupancy value (Qact); and comparing the actual queue occupancy value (Qact) with any one of a minimum threshold (Tmin) and a maximum threshold (Tmax) and providing thereby a result (RES (UP; DO; REM)); and according to the (RES (UP; DO; RET)) re-scaling the actual queue occupancy value (Qact) and the previous drop priority parameter value (DPPprev) and providing thereby, respectively, a re-scaled queue occupancy value (RE-Qact) and the actual drop priority parameter value (DPPact).

The invention provides a medium-short term coordinated dispatching method for ultra-high voltage AC-DC hybrid power grid, and relates to the field of electrical power system power generation dispatching. When electricity is transmitted from west to east, feed-in scale is sharply increased in a low peak period, and load adjustment difficulty of a receiving power network is significantly increased, causing a power grid medium-short term start-up mode and running plan arrangement difficulty to be increased accordingly. The method comprises medium term plan formulation and short term plan formulation. The medium term plan formulation adopts a month as a decision cycle and a day as a calculation period, on the conditions of guaranteeing electrical power system load requirements and operational parameters of a power station and a machine set, monthly electricity is divided into days, and a daily start-up mode of the machine set is optimized; the short term plant forms a 96-point power output plan of daily electricity according to a short-term load prediction result; rolling amendment is conducted on an original medium-term power generating plan according to the latest prediction information of the load and machine set actual electricity completion situation; according to the technical scheme, frequent start-ups and stops of the power generating machine set are avoided, effective engagement and mutual coordination between a medium term and short term power generating plan are achieved.

The invention provides a heating furnace temperature decision-making method based on a billet optimizing heating curve. The method comprises a first step of initialization of a calculation parameter, a second step of calculation segment division of a heating furnace, a third step of position tracking and temperature tracking of all billets in the heating furnace, a fourth step of judging whether the position tracking and the temperature tracking of all the billets in the heating furnace are finished, returning to the third step if the positioning tracking and the temperature tracking are not finished and continuing if the positioning tracking and the temperature tracking are finished, and a fifth step of judging whether a furnace temperature decision-making cycle arrives, returning to the third step if the furnace temperature decision-making cycle does not arrive and conducting the furnace temperature decision-making if the furnace temperature decision-making cycle arrives. The method is simple in correcting algorithm, fast in calculating speed, high in efficiency and beneficial for achieving online control of the heating furnace, saves energy and reduces consumption.

The invention discloses a data transmission method and device. The data transmission method comprises the following steps: in a decision cycle of a data transmission mode, a transmitting end acquires a current data transmission mode; the transmitting end selects a group of antennae thereof according to the current data transmission mode; the transmitting end transmits data according to the current data transmission mode through the antennae. By the invention, spatial correlation requirement of a multi-antenna data transmission mode can be adapted effectively and the performance of a system can be improved.

The invention discloses an intelligent business address selection system. The system comprises a database unit, an information classification and analysis unit and an information query unit, wherein the database unit is used for constructing an address database containing a large amount of addresses, and a business information database which is associated with the address database, stores data which a querier wants to know, and corresponds to the addresses, and is also used for creating an index. According to the invention, first seven links of a general brand extension decision process are combined by synthesizing a large amount of manually-collected information into unified system query, therefore, the data continuity and the data accuracy are ensured, the decision-making period of the querier is greatly shortened and is specifically shortened from two months to one day or one hour. In addition, by virtue of type-in and analysis of business information associated with address information, the business address selection is quick, and controllable factors increases.

The invention discloses a dynamically adjustable emergency transport scheduling plan generation method. The method comprises the steps of 1: uniformly dividing the whole transport scheduling time intoT cycles, and representing a cycle sequence number of a current decision with r, wherein r belongs to {1,...,T}, and an initial value of r is 1; 2: obtaining environmental information of E cycles starting from the first cycle, running an initial plan decision model, generating an E scheduling cycle-oriented initial plan, executing the plan of the first scheduling cycle, and then enabling r to beequal to r+1; 3: collecting actual execution conditions of decisions generated in the cycles from the first cycle to the (r-1)th cycle, and demand update information from the cycle r to the cycle min(r+E-1,T); 4: according to the information collected in the step 3, running an adjustment decision model, generating scheduling decision plans of multiple cycles starting from the current cycle r, butonly executing the plan belonging to the current cycle r; and 5: updating r to be equal to r+1, and if r is less than or equal to T, going to the step 3, or otherwise, ending the scheduling. The method has the advantages that emergency resource scheduling scheme plans are dynamically optimized and adjusted according to information continuous updating results.

An arbitration technique for determining mappings for a switch is described. During a given arbitration decision cycle, an arbitration mechanism maintains, until expiration, a set of mappings from a subset of the input ports to a subset of the output ports of the switch. This set of mappings was determined during an arbitration decision cycle up to K cycles preceding the given arbitration decision cycle. Because the set of mappings are maintained, it is easier for the arbitration mechanism to determine mappings from a remainder of the input ports to the remainder of the output ports without collisions.

The invention discloses a method for reducing the overall overhead of cloud storage in the presence of a data dependence relationship. The method comprises two parts including a data storage decision model and a data storage strategy decision algorithm. In the data storage decision model, a decision variable of the model is a data storage strategy, and comprises non-storage, multi-replica storage and erasure code storage; a constraint condition of the model is that data availability requirements satisfy a given availability index; an optimization target of the model is that the overall overhead of the system is minimized, wherein the data overall overhead can simultaneously comprise storage overhead and calculation overhead; in the calculation overhead, the data generation time is a random variable; an expected calculation method is given in the invention; in the data storage strategy decision algorithm, when new data is generated, the data storage strategy is directly decided; and, when each decision period is ended, the data storage strategy in a communication branch is updated by respectively using a genetic algorithm or a traversal pruning algorithm in manners of taking the communication branch in a dependence relationship graph as a unit and judging whether the node number of the communication branch is greater than a threshold value or not. Compared with the prior art, the method provided by the invention has the advantage that: on the premise that the data availability is ensured, the overall overhead of the system is reduced.

The invention provides an irrigation decision learning method and device, a server and a storage medium. The method and the device can improve the accuracy and scientificity of irrigation decision making and effectively improve the water resource utilization rate. The irrigation decision learning method comprises the following steps: step 1, acquiring environmental parameters of a target irrigation area in a current decision period; step 2, making an irrigation decision based on the acquired environmental parameters and an initialized decision value function; step 3, after the irrigation decision made in the current decision period is executed, evaluating the irrigation decision, obtaining a feedback reward reflecting the reasonable degree of the irrigation decision, and updating a decision value function by utilizing the feedback reward to obtain a decision value function of the next decision period; and step 4, making an irrigation decision of the target irrigation area in the next decision period by utilizing the decision value function of the next decision period and the obtained environmental parameters of the next decision period, and then entering the step 3 to carry out thenext round of irrigation decision learning and making process.

The embodiment of the invention provides a bandwidth control method and device, and relates to the technical field of computers, and the method comprises the steps that edge proxy equipment obtains bandwidth configuration information and bandwidth data of a target channel in a previous decision period; and when it is determined that the decision center device does not generate the center control strategy of the target channel, generating an edge control strategy of the target channel according to the bandwidth configuration information and the bandwidth data of the target channel in the previous decision period, and then sending the edge control strategy of the target channel to the edge resource device. Because the edge proxy device generates the edge control strategy and issues the edgecontrol strategy to the edge resource device when the decision center device does not generate the center control strategy, the method for issuing the edge control strategy before the bandwidth valueof the client channel does not reach the control threshold can effectively solve the problem of center decision hysteresis. Therefore, the sudden increase of the bandwidth is avoided, and the CDN platform and the service of each client can be effectively protected.

The invention provides a side cloud collaborative digital twin intelligent production scheduling application operation position adaptation method. An applied intelligent production scheduling system comprises a cloud layer, an edge layer and a device layer. The generated digital twinning systems are arranged on the edge layer and the cloud layer; and an adaptive strategy controller is arranged on the cloud layer. The adaptation strategy controller dynamically senses edge and cloud resource states, application attributes and the like in the production process, and carries out real-time adaptation switching on the operation position of the digital twin intelligent production scheduling application under two conditions of a determined environment and a non-determined environment by taking minimization of digital twin intelligent production scheduling decision period delay as a target; and real-time adaptive switching is performed on the operation position of the digital twin intelligent production scheduling application based on prediction in an uncertain environment. By adopting the method provided by the invention to determine the operation position of the digital twinning intelligent production scheduling application, the performance loss caused by delay of data interaction and instruction issuing based on a digital twinning system is reduced, and the guarantee based on the digital twinning production scheduling precision is improved.

The embodiment of the invention discloses a data processing method and a data processing device, and the method comprises the steps: obtaining the first bandwidth adjustment information and bandwidth adjustment result of each business flow in M business flows in a first exploration period, and obtaining the bandwidth adjustment result of each business flow according to the first bandwidth adjustment information and the bandwidth adjustment result, determining second bandwidth adjustment information of each service flow in the first decision period; and adjusting each service flow according to the second bandwidth adjustment information; wherein the bandwidth adjustment result is obtained by adjusting each service flow according to the first bandwidth adjustment information; wherein M is an integer greater than or equal to 2. In this way, it is beneficial to determine bandwidth allocation information between multiple traffic flows.

The invention discloses a method and a device for formulating a patrol flight scheme of a helicopter. The method includes the steps of firstly, taking full consideration of a line patrol task, the voyage of the helicopter, a patrol line and peripheral geographic environments and primarily selecting alternative temporary take-off and landing sites suitable for take-off and landing from the line to be patrolled; then listing all combination modes of the temporary take-off and landing sites and enabling each combination mode to serve as one site selection scheme; and finally enabling each site selection scheme to correspond to various line patrol path combination modes, finding one line patrol path combination mode with the shortest flight time as the line patrol flight scheme of each site selection scheme, or finding one site selection scheme with lowest flight cost and the line patrol flight scheme of the site selection scheme, and utilizing the line patrol flight scheme as the optimum line patrol flight scheme. Existing information is sufficiently used, the influence of all factors is comprehensively considered, and a permutation and combination process with the maximum calculated amount is completed by a computer, and accordingly, the optimum line patrol flight scheme can be accurately determined, staff involvement is reduced, decision-making period is shortened, the operating efficiency is improved, and the line patrol cost is reduced.

The invention discloses a business resource allocation method and device, equipment and a storage medium, and the method comprises the steps: obtaining original data related to a resource applicant, and enabling the original data to comprise one or more types of bank data, credit investigation data, tax data, business data and judicial data; processing the original data, and calculating index dataaccording to the processed original data; obtaining a score of a resource applicant according to the index data; performing rating mapping based on the score of the resource applicant to obtain the rating result of the resource applicant; and calculating a resource allocation result based on the rating result of the resource applicant, and outputting the resource allocation result to a resource allocation party, so that the resource allocation party performs resource allocation based on the resource allocation result. The obtained resource allocation result is more comprehensive and objective, the accuracy is higher, and the decision period is greatly shortened, so that the resource application period is greatly shortened.

An arbitration technique for determining mappings for a switch is described. During a given arbitration decision cycle, an arbitration mechanism maintains, until expiration, a set of mappings from a subset of the input ports to a subset of the output ports of the switch. This set of mappings was determined during an arbitration decision cycle up to K cycles preceding the given arbitration decision cycle. Because the set of mappings are maintained, it is easier for the arbitration mechanism to determine mappings from a remainder of the input ports to the remainder of the output ports without collisions.

A video conference system based on Soft-MCU includes a server, a video conference terminal BHU-ICT, a computer terminals and video communication system procedures, wherein, the video conference terminal BHU-ICT is composed by a digital video camera, a microphone, and an audio and video encoder. The digital video camera and the microphone are respectively connected to the audio and video encoder, the encoder, the server and the computer terminal are connected on the network, and the procedures are respectively running in the server and the computer terminal. The digital video camera and the microphone acquire the audio and video information of the meeting place, and use the encoder to execute hardware coding. A media control procedure runs on the computer terminal, and an MCU service procedure run on the server terminal, uses the Ethernet to exchange the audio and video information with the encoder, for performing audio and video communications. All of the texts, requests, commands andother information formats in the system control signaling system conform to the XML standard. The invention constructs a remote network video communications platform with high-quality, high stabilityand high security, which can be widely used in party and government organs and enterprises and institutions. The invention has advantages of reduced system cost, reduced travel costs, shortened decision cycles and enhanced work efficiency.

The invention discloses a city traffic signal optimization control method and belongs to the field of automatic control. The method comprises the following steps: 1, selecting a control scheme, and defining traffic states, an instantaneous crossing income and control motion; 2, a sampling phase employs induction control, and at each sampling time, sending images shot by traffic pick-up heads and phases to a servers; 3, the server extracting the traffic states from the images and calculating the instantaneous crossing income; 4, the server calculating the probability that each discrete traffic state occurs and the crossing income, obtaining an optimal control policy by maximizing an average crossing income and sending the policy to a crossing machine; 5, the crossing machine setting that a control policy of a traffic signal controller is the policy obtained in the fourth step, and setting that a decision period is equal to a sampling period; and 6, the traffic signal controller obtaining the discrete traffic states from the crossing machine, querying the control policy, and sending the control motion to the crossing machine for execution. According to the invention, the average performance index of crossing traffic in a quite long time period can be improved.

The invention discloses an energy strategy management intelligent system adopting multi-object overall optimization. The system comprises a big data collector, a data processor, a subsystem object and constraint generator, an object and constraint screening module, a large system global optimization solving module, a technical and economic evaluation module, and an energy prediction early warning system. The system provided by the invention is based on ideas of the system theory, regards energy, economy, environment, society and technology as a large system, and utilizes the multi-object overall optimization technology, so that integral body, dynamic and non-linear issues planned in the energy-economic system engineering are tackled, the planning and decision-making period is effectively shortened, the coordination cost is reduced, the decision-making process is enabled to be more objective, transparent, stable and reliable, and the system has advantages of being comprehensive, precise, flexible and easy to operate and becomes a new tool for carrying out strategic plans in energy engineering management.

The embodiment of the present invention discloses a method for selecting and judging a remote radio frequency unit in the downlink direction, and the embodiment of the invention also discloses a device for selecting and judging a remote radio frequency unit in the downlink direction. Wherein, the method includes: the RRU selection decision period of the current RRU common cell arrives; according to the current downlink measurement amount of the user terminal and the uplink measurement amount of each RRU, it is determined that at least one RRU in the RRUs transmits the downlink carrier signal, or, According to the current downlink measurement amount of the user terminal, the uplink measurement amount of each RRU, and the power specification of each RRU, it is determined that at least one of the RRUs transmits a downlink carrier signal, and the downlink channel includes: a traffic channel and an independent dedicated control channel. By adopting the present invention, an appropriate RRU can be selected in good time to transmit a downlink carrier signal, and a reasonable conversion of the RRU transmission mode among single transmission, double transmission, and multiple transmission modes can be realized; the probability of RRU multi-transmission or double transmission can be reduced, thereby reducing network interference and RRU energy consumption.

The invention discloses an improved distributed multi-agent system that can be used for micro-grid multi-target energy management, including: at least one micro-grid unit; a main container, which includes a basic Agent, an aggregated electrical Agent and an Agent platform; the basic Agent is used for Making decisions, negotiating and executing microgrid unit solutions, its running time is defined as the decision cycle, which starts asynchronously with unpredictable events, and the completion time is dynamically adjusted; the basic Agent also includes an allocation unit, and the microgrid units are divided according to functions , and the allocation unit is paired with at least one type of basic Agent; at least one Agent container, each Agent container includes a unit electrical Agent and at least one basic Agent, the unit electrical Agent is bound to a microgrid unit, and sends the corresponding electrical information to Aggregate electrical Agent; each basic Agent communicates with Agent protocol. Establish communication among agents, make decisions on their own behavior based on their own information and local environmental information, and finally achieve the overall optimization control goal of the microgrid.

The embodiment of the invention discloses a multi-element main body hardware management method and a related product. The method comprises the following steps: a server sends a decision inquiry page about a decision issue to each of a plurality of authorization clients, wherein the decision inquiry page is used for collecting identity information and decision options of a decision subject corresponding to each authorization client; the server receives a decision feedback page sent by each authorization client in the plurality of authorization clients to obtain a plurality of decision feedbackpages; the server determines a decision result according to the plurality of decision feedback pages; the server extracts decision suggestions from the plurality of decision feedback pages; and the server sends the decision result and the decision suggestion to each authorization client in the plurality of authorization clients. According to the scheme, the decision period is greatly shortened, and the decision efficiency is improved.

The invention provides a method for adapting the running position of an edge-cloud collaborative digital twin intelligent production scheduling application. The applied intelligent production scheduling system includes a cloud layer, an edge layer and a device layer. The generated digital twin system is set at both the edge layer and the cloud layer; an adaptation strategy controller is set at the cloud layer. The adaptation strategy controller dynamically senses the edge and cloud resource status, application attributes, etc. in the production process, aiming at minimizing the delay of the digital twin intelligent production scheduling decision cycle, and has two types of intelligence for the digital twin: deterministic environment and non-deterministic environment. Real-time adaptation and switching of the running position of the production scheduling application; real-time adaptation and switching of the running position of the digital twin intelligent production scheduling application based on prediction in a non-deterministic environment. Using the method of the invention to determine the running position of the digital twin intelligent production scheduling application reduces the performance impairment caused by the digital twin based system due to the delay in data interaction and command issuance, and improves the guarantee of the digital twin based production scheduling accuracy.

The invention discloses a dual-pressure transmitter redundancy design algorithm based on air pressure prediction. The dual-pressure transmitter redundancy design algorithm comprises a pressure judgment algorithm, an air pressure prediction algorithm and a pressure decision algorithm. According to the pressure judgment algorithm, the output pressure of the two sets of pressure transmitters is collected for multiple times in a decision-making period. Then, pressure range judgment, pressure trend judgment, multi-point average operation and effective pressure point number judgment are carried out for judging whether the pressure acquired by the pressure transmitter is effective or not, and effective pressure values P1 and P2 are given out; according to the air pressure prediction algorithm, a predicted pressure value Py after a decision period is calculated according to the received inflation time tc, the accumulated exhaust time tp, the inflation amount and the exhaust amount; according to the pressure decision algorithm, effective pressure values P1 and P2 output by the pressure judgment algorithm and a predicted pressure value Py output by the air pressure prediction algorithm are received, and system decision pressure P is output; compared with a traditional air pressure control method, the algorithm adopts two sets of pressure transmitters to test pressure in a redundant mode, and the test precision and reliability can be greatly improved.

The invention discloses a method, device, setting and storage medium for allocating business resources. By obtaining original data related to resource applicants, the original data includes bank data, credit data, tax data, industrial and commercial data and judicial data. One or more of the data; process the original data, and calculate index data according to the processed original data; obtain the resource applicant’s score according to the index data; Rating mapping, obtaining the rating result of the resource applicant; calculating the resource allocation result based on the rating result of the resource applicant, and outputting the resource allocation result to the resource allocator, so that the resource allocator can allocate resources based on The result is resource allocation; the obtained resource allocation results are more comprehensive, objective, and more accurate, and the decision-making cycle is greatly shortened, which greatly reduces the resource application cycle.

The invention relates to the technical field of power system automation, in particular to a power and energy balance optimization method based on wind power capacity credibility. The power and energybalance optimization method comprises the steps that data of thermal power and hydropower units of provinces in a study area are collected; load forecast curves and wind power forecast outputs of theprovinces in a decision-making period are obtained; wind power credible forecast outputs are obtained according to the wind power capacity credibility and the wind power forecast outputs; a tie line transmission power plan is drawn up, the maximum and the minimum of transmission power are set in time segments; the minimum operating cost of the whole network is used as a target operation function,and power balance constraint, unit operation constraint, tie line transmission constraint, electric quantity constraint and rotation reserve constraint are used as constraint conditions, so that a power and energy balance optimization model is established; and the optimal outputs of generator sets of the provinces and the transmission power of tie lines between the provinces are obtained by solving the model. The uncertain wind power outputs are converted into credible unit outputs through wind power credibility calculation, and the problem that a traditional power and energy balance method cannot properly consider wind power is solved.

The invention provides an air combat maneuver decision-making method and device, and the method comprises the steps: constructing an aircraft air combat model according to the actual performance limiting factors of an aircraft; respectively setting different decision periods for different aircraft states according to the dynamic response time of the aircraft states under the control of a flight control system; according to the aircraft states after the decision period is set, determining target air combat maneuvering instructions corresponding to different aircraft states; and according to the target air combat maneuvering instruction and the aircraft air combat model, performing an air combat maneuvering decision on the aircraft, wherein the aircraft states comprise the height, the course angle and the speed. According to the air combat maneuver decision-making method and device, different decision-making periods are set for maneuver decisions of different aircraft states through the dynamic response characteristics of different aircraft states under the control system, the contradiction between the real-time performance and the accuracy of air combat maneuver decision-making can be effectively relieved, and rapid and efficient air combat maneuver decision making is realized.

The invention discloses a method and device for determining agent processing, equipment and a storage medium, and belongs to the technical field of artificial intelligence. The method comprises the steps of determining a first environment feature corresponding to a target agent every time a decision period is reached; acquiring a second environment feature corresponding to a neighbor agent of the target agent, wherein the neighbor agent is a preset agent communicating with the target agent in a multi-agent system; generating a target fusion environment feature based on the first environment feature and the second environment feature; and inputting the target fusion environment features into a strategy network to obtain target processing which needs to be executed by the target agent in the current decision period. By adopting the method and the device, the efficiency of the intelligent agent for determining the processing required to be executed in the decision period can be improved.