A web point intelligent service method and device, electronic equipment and storage medium

Abstract

The application discloses a kind of network point intelligent service method, device, electronic equipment and storage medium.Method according to the state space of decision model in intelligent agent, at least one state information of target network point belonging to state space is obtained;State space includes at least one state as follows: historical passenger flow state, predicted passenger flow state of future specified period, window service state, service provider state and each internet of things equipment state;At least one state information is input into decision model, to output at least one action information belonging to the action space of decision model;Action space includes at least one action as follows: window service adjustment action, internet of things equipment linkage control action and service provider guiding action;According to at least one action information, corresponding action instruction is sent to the controlled object of action information.The technical scheme of the present application improves the overall collaborative service efficiency of network point resources.

Description

Technical Field

[0001] This invention relates to the field of service outlet technology, and in particular to a smart service outlet method, device, electronic device, and storage medium. Background Technology

[0002] Traditional bank branch services primarily rely on customers taking numbers, waiting in line, and receiving manual guidance. While some branches have introduced queuing systems and self-service terminals, service processes remain relatively fixed, and resource allocation (windows, personnel, equipment) depends heavily on the experience of on-site management personnel, resulting in delayed responses. Customer identification is typically conducted at the counter, lacking proactive anticipation and service guidance. Furthermore, branch resources operate in isolation with insufficient coordination, leading to overall low efficiency in collaborative service. Summary of the Invention

[0003] This invention provides a method, device, electronic device, and storage medium for intelligent services at service outlets, in order to improve the overall collaborative service efficiency of service outlet resources.

[0004] According to one aspect of the present invention, a method for providing intelligent services at service outlets is provided, the method comprising:

[0005] Based on the state space of the decision model in the intelligent agent, obtain at least one state information of the target outlet belonging to the state space; wherein, the state space includes at least one of the following states: historical passenger flow state, predicted passenger flow state for a specified future period, window service state, service provider state, and the state of each IoT device;

[0006] At least one of the aforementioned state information is input into the decision model to output at least one action information belonging to the action space of the decision model; wherein, the action space includes at least one of the following actions: window service adjustment action, IoT device linkage control action, and service provider guidance action;

[0007] Based on at least one of the action information, a corresponding action instruction is sent to the controlled object of the action information.

[0008] According to another aspect of the present invention, a smart service device for service outlets is provided, the device comprising:

[0009] The network information acquisition module is used to acquire at least one state information of the target network belonging to the state space based on the state space of the decision model in the intelligent agent; wherein, the state space includes at least one of the following states: historical passenger flow status, predicted passenger flow status for a specified future period, window service status, service provider status, and status of each Internet of Things device;

[0010] An action output module is used to input at least one of the aforementioned state information into the decision model, so as to output at least one action information belonging to the action space of the decision model; wherein, the action space includes at least one of the following actions: window service adjustment action, IoT device linkage control action, and service provider guidance action;

[0011] An action instruction sending module is used to send a corresponding action instruction to the controlled object of the action information based on at least one of the action information.

[0012] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0013] At least one processor; and

[0014] A memory communicatively connected to the at least one processor; wherein,

[0015] The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the intelligent branch service method according to any embodiment of the present invention.

[0016] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the intelligent branch service method according to any embodiment of the present invention.

[0017] The technical solution of this invention obtains at least one state information belonging to the state space of the decision model in the intelligent agent. The state space includes at least one of the following states: historical passenger flow status, predicted passenger flow status for a specified future period, window service status, service provider status, and status of each IoT device. At least one state information is input into the decision model to output at least one action information belonging to the action space of the decision model. The action space includes at least one of the following actions: window service adjustment action, IoT device linkage control action, and service provider guidance action. Based on at least one action information, corresponding action instructions are sent to the controlled object of the action information. By adopting a technical means of collaboratively scheduling overall network resources and providing proactive guidance actions, this solution addresses the problem of existing networks lacking proactive prediction and service guidance, and the isolated operation and insufficient collaboration of network resources, resulting in low overall collaborative service efficiency of network resources. This improves the overall collaborative service efficiency of network resources.

[0018] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1a A flowchart of a smart service method for service outlets provided in Embodiment 1 of the present invention;

[0021] Figure 1b A flowchart of another intelligent service method for service outlets provided in Embodiment 1 of the present invention;

[0022] Figure 2 This is a schematic diagram of the structure of an intelligent service device for service outlets provided in Embodiment 2 of the present invention;

[0023] Figure 3 This is a schematic diagram of the structure of an electronic device that implements the intelligent service method for service outlets according to embodiments of the present invention. Detailed Implementation

[0024] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0025] It should be noted that the terms "comprising" and "having" and any variations thereof in the specification, claims and accompanying drawings of this invention are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to such processes, methods, products or devices.

[0026] Example 1

[0027] Figure 1a This is a flowchart of a smart service method for service outlets provided in Embodiment 1 of the present invention. This embodiment is applicable to situations where smart services are provided within service outlets. The method can be executed by a smart service device for the outlet, which can be implemented in hardware and / or software. This smart service device can be configured in a server integrated with an intelligent agent. For example... Figure 1aAs shown, the method includes:

[0028] S110. Based on the state space of the decision model in the intelligent agent, obtain at least one state information of the target outlet belonging to the state space; wherein, the state space includes at least one of the following states: historical passenger flow state, predicted passenger flow state for a specified future period, window service state, service provider state, and the state of each IoT device.

[0029] Historical passenger flow status refers to passenger flow information of the target outlet during historical business hours. Predicted passenger flow status for a specified future period refers to future passenger flow (e.g., passenger flow status in the next 30 minutes) predicted based on historical passenger flow status using a predictive model (e.g., Long Short-Term Memory network). Window service status may include at least one of the following: the type of service currently being processed, the estimated remaining processing time, and the window's availability status (e.g., open, closed, or busy level). Service provider status includes at least one of the following: service provider location, service provider skills, and service provider availability. IoT device status includes at least one of the following: form-filling station availability, seat occupancy status, and security glass privacy status.

[0030] In this embodiment, a simulation environment for a target branch can be constructed, and key elements of a real-world scenario can be simulated within this environment. For example, a customer flow generator can be used to generate a virtual customer arrival sequence (equivalent to a predicted customer flow state for a specified future time period) based on historical data (such as peak passenger flow distribution and business type ratios). A resource status simulator can be used to simulate window status (open, closed, or busy), employee status (such as employee location and skills), and IoT device status (such as seat occupancy status, form-filling station idle status, and security glass privacy status). Time consumption models for different business types can be defined (such as a normal distribution to simulate processing time).

[0031] In one alternative implementation, the agent's decision model can be obtained by: constructing the state space and action space of the decision model; inputting the state information from the training samples into the decision model to output the action information determined from the action space; evaluating the action information according to the reward function, and feeding back the evaluation results to train the decision model.

[0032] The decision-making model in this embodiment can be a model built based on reinforcement learning. Therefore, this embodiment can define the state space and action space of the agent based on the above-mentioned branch simulation environment. Specifically, the state space can refer to the set of all states that the agent may be in in the branch simulation environment, which can describe all environmental conditions that the agent can perceive. The action space can refer to the set of all possible actions that the agent can perform in a specific state. The action space can be discrete or continuous scheduling operations. Example of discrete action: Switch window 1 to business service, close window 2, dispatch employee A to area X, activate self-service terminal guidance task. Example of continuous action: Adjust the service weight of each window (e.g., allocate 70% of resources to the business service window to handle VIP business).

[0033] The reward function in this embodiment can be constructed by weighted summation of multidimensional indicators, which may include at least one of the following: waiting time reduction, customer satisfaction, employee workload balancing, and resource switching costs. For example, the reward function can be expressed as Reward = w1. (Reduction in waiting time) + w2 (Customer Satisfaction) +w3 (Employee workload balance) -w4 (Resource switching cost) can be evaluated based on the reward function to obtain positive and negative rewards. For example, a positive reward can be obtained for shortening customer waiting time, providing timely service to high-anxiety customers, and having a window utilization rate of >90%. A negative reward can be obtained for frequent window switching, excessive distance traveled by employees across areas, and simple business occupying manual windows.

[0034] Based on the above optional implementation methods, after evaluating the action information according to the reward function and before feeding back the evaluation results to train the decision model, it may further include: comparing the reward with a preset reward distribution range; and filtering out rewards that exceed the preset reward distribution range based on the comparison results.

[0035] This reward filtering setting is intended to prune rewards, limit extreme rewards, and prevent instability in the training of the decision model.

[0036] After training the decision-making model based on the evaluation results, the process may also include: conducting an offline comparison between the decision-making model's decision-making strategy and the benchmark rules; and manually adjusting the weights of various indicators in the reward function based on the offline comparison results to ensure that the decision-making strategy aligns with business priorities.

[0037] This embodiment can compare the decision-making strategy and benchmark rules of the decision-making model offline in the branch simulation environment. For example, the average waiting time decreases by 40% and the window utilization rate increases to 92%. Then, based on the offline comparison results, the weights of each indicator (w1, w2, w3, w4) in the reward function are manually adjusted to ensure that the decision-making strategy is in line with business priorities.

[0038] This embodiment deploys the trained decision-making model to real target locations to respond to location status in real time and output action information. Specifically, it can obtain at least one piece of real-time status information of the target location based on a predefined state space, such as real-time customer flow status, real-time window service status, real-time employee (i.e., service provider) status, and real-time IoT device status.

[0039] S120. Input at least one state information into the decision model to output at least one action information belonging to the action space of the decision model; wherein, the action space includes at least one of the following actions: window service adjustment action, IoT device linkage control action, and service provider guidance action.

[0040] At least one action information output by the decision model in this embodiment can be based on action constraints. These action constraints can be based on business rule verification and security interception definitions, such as preventing switching from a corporate transaction window to a personal window if there is currently a queue, and prohibiting high-risk actions (e.g., closing all cash transaction windows).

[0041] S130. Based on at least one action information, send the corresponding action instruction to the controlled object of the action information.

[0042] This embodiment can send action information to the corresponding actual executor within the target network.

[0043] In one optional implementation, sending a corresponding action instruction to the controlled object of the action information based on at least one action information may include: sending a window service adjustment instruction to the queuing system based on a window service adjustment action; sending a device control instruction to the IoT device collaborative control network based on an IoT device linkage control action; and sending a service instruction to the smart wearable device worn by the target service provider based on a service provider guidance action.

[0044] In this embodiment, the window service adjustment action can be controlled and executed by the queuing system, and the corresponding instruction is, for example, to change window Y from personal service to corporate service; the IoT device linkage control action can be executed by the IoT device collaborative control network to control the IoT devices, and the corresponding instruction is, for example, to trigger the form filling station X to display the account opening form; the service provider guidance action can be instructed by smart wearable devices (such as mobile terminals or smart glasses, etc.) to employees, and the corresponding instruction is, for example, to dispatch an account manager to area A to receive customer 1.

[0045] Optionally, the state space may also include user characteristics of users within the branch. Accordingly, the branch intelligent service method of this embodiment can be referred to Figure 1b ,include:

[0046] S1101. Based on the state space of the decision model in the intelligent agent, obtain the user characteristics of the current user entering the current branch, the current real-time passenger flow distribution, the current status of each IoT device, the current business status of each window, and the current status of each service provider.

[0047] Optionally, obtaining the user characteristics of the current user entering the current branch may include: receiving the identification information of the current user reported by the staff at the branch entrance, and calling a multimodal perception system to obtain the multimodal biometrics of the current user; the multimodal biometrics include the current user's gait characteristics and / or voice characteristics; calling the cloud platform based on the identification information to obtain the current user's historical business information; comprehensively analyzing the current user's multimodal biometrics and historical business information to obtain the current user's user characteristics, which include identification information, historical preferences, potential business needs, real-time status index, and business priority. Among these, business priority can be determined through a comprehensive evaluation of multiple factors, including business value contribution, criticality and urgency of business type, business complexity, user status, historical risks, and branch global resource load balancing.

[0048] In this embodiment, user data can be collected through an authorization mechanism. Specifically, 1) a quick authorization channel: customers can pre-sign the "Authorization Form for the Use of Biometrics and Historical Data" through a bank application, and the authorization status is synchronized to the edge computing node; 2) an on-site authorization channel: customers sign the authorization form at the entrance through a smart terminal, clearly stating the scope, purpose, data type, and storage period of the data collection. The "Authorization Form for the Use of Biometrics and Historical Data" can break down "Biometric Collection," "Cross-Branch Data Sharing," and "Automated Tagging and Decision Making" into independent, explicitly selectable authorization items, providing clear and easy-to-understand explanations. If a customer does not agree to any of these items, they will refuse to sign the authorization form.

[0049] In addition to signing the "Authorization Agreement for the Use of Biometric and Historical Data," the following measures can also be taken: 1) Strengthen the informed consent system by separating biometric collection, cross-site data sharing, and automated decision-making into independent authorization items, supplemented with explanations, and prominently displaying the scope of the contactless identification area and the content to be collected at the site entrance; 2) Implement the principle of minimizing data processing, prioritizing edge processing of biometric data and uploading only the identification results, and setting a short-term storage strategy; strictly limit the scenarios for status analysis and provide an option to close it in real time; 3) Protect user rights by providing convenient channels for customers to view, correct, or delete tags, rejecting automated decision-making, and adding anti-spy screens and operating procedures to smart wearable devices; 4) Enhance data security by encrypting biometric data throughout the process and using dedicated secure storage, and overlaying access control token technology for cross-site transmission; 5) Introduce third-party institutions to conduct privacy impact assessments and algorithm ethics audits, establish a manual review mechanism for high-risk decisions, ensure non-discrimination, and conduct regular reviews.

[0050] The multimodal perception system in this embodiment may include devices such as millimeter-wave radar and microphone arrays. After a user enters the target location, millimeter-wave radar can be used to assist in acquiring gait features, enhancing recognition accuracy; microphone arrays can be used to capture voice segments to assist in verification.

[0051] Furthermore, based on the user identifier, the user's historical business information can be retrieved, and by comprehensively analyzing the user's multimodal biometrics and historical business information, the user characteristics of the current user can be determined.

[0052] In one scenario, for users who have already conducted business at other branches of the same company and who meet the company's data sharing policy, cross-branch historical business information can be retrieved through the cloud platform. Cross-branch data transmission can be encrypted using the national cryptographic standard SM4 and stored on the blockchain to ensure security and traceability.

[0053] In this embodiment, the criticality of a business type can be determined based on its high-value or high-risk attributes. For example, large-amount financial advisory services and complex loan applications have high-value attributes and require in-depth involvement from senior account managers, allowing for the allocation of corresponding resources. Large-amount cash withdrawals, password resets, and card loss reporting and unblocking services involve customer fund security and compliance requirements, thus having high-risk attributes and requiring rapid, accurate, and secure processing. To avoid customer anxiety or risk events, these services have a higher priority.

[0054] The urgency of a service type can be determined based on the user's potential business needs and real-time behavior analysis. For example, behaviors such as a user heading straight to the ATM area upon entering the branch (strongly implying an urgent need for withdrawal), frequently checking their watch / phone, or fidgeting in the waiting area can indicate that the user's business is urgent.

[0055] Business complexity can be analyzed based on the processing time of similar historical transactions. For example, for transactions expected to be short-term and simple (such as inquiries and small deposits / withdrawals), even if the customer has an urgent request, the system may be more inclined to guide them to self-service terminals or fast-track windows, freeing up human resources at manual counters. For transactions expected to be long-term and complex (such as account opening and complex financial management contracts), even if the customer is currently in a stable state, they should be assigned to a suitable (skilled) human counter as early as possible to avoid backlogs later.

[0056] Historical risks can be analyzed based on users' past behavior. For example, customers who have had poor service experiences or given negative feedback in the past may be given higher priority in this service to improve customer relationships and prevent recurrence of problems.

[0057] Global resource load balancing at a branch network refers to balancing the load of various service resources across the entire branch network. For example, when allocating resources to a high-priority service (such as a dedicated account manager), the impact of this decision on the load of other waiting customers and staff can be considered. For instance, during extremely busy periods, even for VIP customers, staff can be advised to offer brief reassurance and inform them of the estimated wait time, rather than immediately interrupting current services.

[0058] S1102. Input user characteristics, current real-time passenger flow distribution, current status of each IoT device, current status of each window service, and current status of each service provider into the decision model to generate window service adjustment instructions, device control instructions, and service instructions.

[0059] The decision-making model in this embodiment can be fine-tuned online. For example, if the actual reward deviates significantly from the expected reward during the operation of the decision-making model at the target site (e.g., due to a sudden surge in customer traffic), a lightweight retraining can be initiated. The retrained new strategy can then be tested in an isolated site simulation environment, and deployed to the actual target site only after confirming its effectiveness. For example, if actual employees flag abnormal decisions made by the decision-making model (e.g., a high-risk business has too low a priority), the reward function can be corrected accordingly.

[0060] S1103. Based on the window service adjustment action, send the window service adjustment instruction to the queuing system.

[0061] S1104. Based on the IoT device linkage control actions, send device control commands to the IoT device collaborative control network.

[0062] The IoT device collaborative control network in this embodiment can communicate with each IoT device based on a dedicated communication protocol for IoT device group control. This protocol can use CRC16 checksum to ensure data integrity and can implement a command retransmission mechanism to cope with network fluctuations or packet loss.

[0063] For example, when the system detects that a user is approaching a specific form-filling station, it can trigger the touchscreen to automatically display the user's frequently used form templates via this protocol; the seat has a built-in pressure sensor, and when it detects that a user is sitting down, it reports the event via this protocol, which can be used to fine-tune the queuing strategy (such as extending the waiting tolerance for customers who are currently processing business); when a user is processing a sensitive business involving privacy, the protocol can be used to control the glass to switch to a frosted state to protect privacy.

[0064] S1105. Based on the actions guided by the service provider, send service instructions for the current user to the smart wearable device worn by the target service provider.

[0065] To enable those skilled in the art to better understand the intelligent service method for service outlets in this embodiment, the following is an application scenario provided by this embodiment.

[0066] As Mr. Zhang stepped into the branch entrance:

[0067] 1) Report Mr. Zhang's identity

[0068] The staff at the entrance checked Mr. Zhang's documents and reported his identity.

[0069] 2) Retrieve historical data

[0070] Real-time retrieval of Mr. Zhang's historical data (via edge computing nodes, latency <100ms): VIP level L5, 6 large-amount financial consultations in the past 3 months, risk preference "conservative", no complaint records.

[0071] 3) Real-time behavior analysis and user information

[0072] Millimeter-wave radar tracked Mr. Zhang's movement path and showed that he went straight to the financial consultation area (not the waiting area) after entering. Analysis showed that Mr. Zhang's business was urgent financial consultation.

[0073] Mr. Zhang's user information can be:

[0074] Customer ID: CN123456;

[0075] Name: Mr. Zhang;

[0076] VIP Level: L5;

[0077] "Historical Business": ["Large-sum financial advisory services"];

[0078] "Potential Demand Tags": ["Urgent Financial Consultation", "Time-Sensitive"];

[0079] "Business Priority": "High (VIP + Urgent)";

[0080] Special Note: Priority service required.

[0081] 4) Obtaining global status of network points and outputting decision-making data.

[0082] The decision-making model can determine the most suitable employee (employee ID: EMP007, with financial management qualifications, currently located 5 meters away from Mr. Zhang) based on the above user information and the overall status of the branch (real-time customer flow, real-time window business status, real-time employee status, real-time IoT device status, etc.).

[0083] The smart wearable device (such as smart glasses) worn by Manager Li can receive the following information and overlay it visually: In the real field of vision, a virtual identifier appears above Mr. Zhang's head: a gold border (VIP logo) + a lightning bolt icon (urgent) + the text label "VIP - Urgent Financial Management". Service strategy suggestion: Dynamically generate the following in the sidebar: 1. Immediately guide to the VIP room (prioritize customers in the queue); 2. Handle the matter personally (do not transfer to the lobby assistant); 3. Enable privacy protection (linked to the frosted glass).

[0084] If Mr. Zhang moves at a steady pace: We can suggest "guiding him to the wealth management area to queue, with an estimated wait of 5 minutes," using a standard greeting. If Mr. Zhang is a first-time customer: We can suggest "assistance with filling out forms and guidance to the self-service terminal," focusing on explaining the process (e.g., if it's his first time, he can first learn about the procedure).

[0085] In the above scenario, the input state vector of the decision-making model can be [current time 09:30, customer flow density in the financial management area: 0.8, window waiting time: [5, 10, 8] minutes, employee EMP007: idle / skilled financial management, LSTM prediction of peak customer flow at 09:45]. The decision action can be: dispatch EMP007 to guide Mr. Zhang to a certain room, dynamically allocate window 3 as financial management priority.

[0086] The decision-making action instruction can be issued to the queuing system and employee terminals, allowing Mr. Zhang to enter the VIP room with zero waiting time. Record Mr. Zhang's actual waiting time as 0 minutes and the positive reward function as +8. If the actual reward is lower than expected (e.g., employee response delay), a lightweight retraining process (<1 minute) can be initiated.

[0087] When Mr. Zhang was guided to a certain room for financial consultation, the workflow of the IoT device collaborative control network can be as follows:

[0088] The control center generates a command packet: [Command ID: CMD_GLASS_FOG, Device ID: DEV_VIP01, Parameters: Atomization Level 3, CRC16: 0x3A7B]; calculates the CRC16 checksum of the command data (e.g., 0x8005); the command packet is sent to the edge gateway via the IoT device collaborative control protocol (lightweight UDP custom protocol). The edge gateway forwards the command to the VIP room bulletproof glass controller; after receiving the command, the controller verifies the CRC16 checksum. If the verification is successful, bulletproof glass atomization can be executed, and an ACK (Acknowledge character) confirmation packet is returned. If the verification fails (verification error or packet loss), the command can be discarded without response. Waiting for ACK (timeout set to 500ms): If an ACK packet is received within the set time, the process can be considered complete (atomization effective). If no ACK packet is received within the timeout, a retransmission mechanism can be triggered. The first retransmission is delayed by 100ms, and a maximum of 3 retries are allowed. If 3 attempts fail, the device is marked as faulty, and an alarm is issued to the maintenance personnel.

[0089] The technical solution of this embodiment obtains at least one state information belonging to the state space of the decision model in the intelligent agent. The state space includes at least one of the following states: historical passenger flow status, predicted passenger flow status for a specified future period, window service status, service provider status, and status of each IoT device. The at least one state information is input into the decision model to output at least one action information belonging to the action space of the decision model. The action space includes at least one of the following actions: window service adjustment action, IoT device linkage control action, and service provider guidance action. Based on the at least one action information, the corresponding action instruction is sent to the controlled object of the action information. By adopting the technical means of collaboratively scheduling the overall network resources and providing proactive guidance actions, this solution solves the problem that existing network resources lack proactive prediction and service guidance, and that network resources operate in isolation with insufficient collaboration, resulting in low overall collaborative service efficiency of network resources. This improves the overall collaborative service efficiency of network resources.

[0090] Example 2

[0091] Figure 2 This is a schematic diagram of the structure of a smart service device for service outlets provided in Embodiment 2 of the present invention. Figure 2 As shown, the device includes: a branch information acquisition module 210, an action output module 220, and an action command sending module 230. Wherein:

[0092] The network information acquisition module 210 is used to acquire at least one state information of the target network belonging to the state space based on the state space of the decision model in the intelligent agent; wherein, the state space includes at least one of the following states: historical passenger flow status, predicted passenger flow status for a specified future period, window service status, service provider status, and status of each Internet of Things device.

[0093] Action output module 220 is used to input at least one of the aforementioned state information into the decision model, so as to output at least one action information belonging to the action space of the decision model; wherein, the action space includes at least one of the following actions: window service adjustment action, IoT device linkage control action, and service provider guidance action;

[0094] The action instruction sending module 230 is used to send a corresponding action instruction to the controlled object of the action information based on at least one of the action information.

[0095] The technical solution of this embodiment obtains at least one state information belonging to the state space of the decision model in the intelligent agent. The state space includes at least one of the following states: historical passenger flow status, predicted passenger flow status for a specified future period, window service status, service provider status, and status of each IoT device. The at least one state information is input into the decision model to output at least one action information belonging to the action space of the decision model. The action space includes at least one of the following actions: window service adjustment action, IoT device linkage control action, and service provider guidance action. Based on the at least one action information, the corresponding action instruction is sent to the controlled object of the action information. By adopting the technical means of collaboratively scheduling the overall network resources and providing proactive guidance actions, this solution solves the problem that existing network resources lack proactive prediction and service guidance, and that network resources operate in isolation with insufficient collaboration, resulting in low overall collaborative service efficiency of network resources. This improves the overall collaborative service efficiency of network resources.

[0096] Optionally, the intelligent service device for service outlets further includes a decision model acquisition module, used for:

[0097] Construct the state space and action space of the decision model;

[0098] The state information from the training samples is input into the decision model to output action information determined from the action space;

[0099] The action information is evaluated based on the reward function, and the decision model is trained based on the evaluation results.

[0100] Optionally, the reward function is constructed by weighted summation of multidimensional indicators, which include at least one of waiting time reduction, customer satisfaction, employee workload balancing, and resource switching costs.

[0101] Optionally, the intelligent service device for service outlets further includes a reward screening module, used to evaluate the action information according to the reward function and before training the decision model based on the evaluation results:

[0102] Compare the reward with a preset reward distribution range;

[0103] Rewards that exceed the preset reward distribution range will be filtered out based on the comparison results;

[0104] Optionally, the intelligent service device for service outlets further includes a weight adjustment module, used after training the decision model based on the evaluation results:

[0105] The decision-making strategy of the decision-making model and the benchmark rule are compared offline.

[0106] Based on the offline comparison results, the weights of each indicator in the reward function are manually adjusted to ensure that the decision-making strategy aligns with business priorities.

[0107] Optionally, the window service status includes at least one of the following: the type of service currently being processed, the estimated remaining processing time, and the window's idle status;

[0108] The service provider status includes at least one of the following: service provider location, service provider skills, and service provider idle status;

[0109] The status of each IoT device includes at least one of the following: idle status of the form filler, occupied status of the seat, and privacy status of the safety glass.

[0110] Optionally, the action command sending module 230 can be used for:

[0111] Based on the window service adjustment action, send a window service adjustment instruction to the queuing system;

[0112] Based on the IoT device linkage control actions, send device control commands to the IoT device collaborative control network;

[0113] Based on the actions guided by the service provider, a service instruction is sent to the smart wearable device worn by the target service provider.

[0114] Optionally, the state space may also include user characteristics of users within the network.

[0115] Correspondingly, the branch information acquisition module 210 can be specifically used for:

[0116] Get the user characteristics of the current user entering the current branch, the current real-time customer flow distribution, the current status of each IoT device, the current status of each window service, and the current status of each service provider;

[0117] The motion output module 220 can be used specifically for:

[0118] The user characteristics, the current real-time passenger flow distribution, the current status of each IoT device, the current status of each window service, and the current status of each service provider are input into the decision model to generate window service adjustment instructions, device control instructions, and service instructions.

[0119] Action command sending module 230 can be used for:

[0120] Based on the window service adjustment action, send a window service adjustment instruction to the queuing system;

[0121] Based on the IoT device linkage control actions, send device control commands to the IoT device collaborative control network;

[0122] Based on the actions guided by the service provider, a service instruction for the current user is sent to the smart wearable device worn by the target service provider.

[0123] Optionally, the branch information acquisition module 210 can also be used for:

[0124] The system receives the identification information of the current user reported by the staff at the branch entrance and calls the multimodal perception system to obtain the multimodal biometrics of the current user; the multimodal biometrics include the current user's gait characteristics and / or voice characteristics.

[0125] The cloud platform is invoked based on the identification information to obtain the current user's historical business information;

[0126] By comprehensively analyzing the multimodal biometrics of the current user and the historical business information, the user characteristics of the current user are obtained. The user characteristics include the identification information, historical preferences, potential business needs, real-time status index, and business priority. The business priority is determined by a comprehensive evaluation of multiple factors, including business value contribution, criticality and urgency of business type, business complexity, user status, historical risks, and global resource load balancing of the network.

[0127] The intelligent service device for service outlets provided in the embodiments of the present invention can execute the intelligent service method for service outlets provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of executing the method.

[0128] Example 3

[0129] Figure 3 A schematic diagram of an electronic device 300 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers or various forms of mobile devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0130] like Figure 3 As shown, the electronic device 300 includes at least one processor 301 and a memory, such as a read-only memory (ROM) 302 or a random access memory (RAM) 303, communicatively connected to the at least one processor 301. The memory stores computer programs executable by the at least one processor. The processor 301 can perform various appropriate actions and processes based on the computer program stored in the ROM 302 or loaded into the RAM 303 from storage unit 308. The RAM 303 can also store various programs and data required for the operation of the electronic device 300. The processor 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0131] Multiple components in electronic device 300 are connected to I / O interface 305, including: input unit 306, such as keyboard, mouse, etc.; output unit 307, such as various types of displays, speakers, etc.; storage unit 308, such as disk, optical disk, etc.; and communication unit 309, such as network card, modem, wireless transceiver, etc. Communication unit 309 allows electronic device 300 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0132] Processor 301 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 301 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 301 performs the various methods and processes described above, such as the branch intelligent service method.

[0133] In some embodiments, the branch office intelligent service method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 308. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 300 via ROM 302 and / or communication unit 309. When the computer program is loaded into RAM 303 and executed by processor 301, one or more steps of the branch office intelligent service method described above may be performed. Alternatively, in other embodiments, processor 301 may be configured to perform the branch office intelligent service method by any other suitable means (e.g., by means of firmware).

[0134] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0135] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0136] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0137] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0138] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0139] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0140] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0141] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for providing intelligent services at service outlets, characterized in that, include: Based on the state space of the decision model in the intelligent agent, obtain at least one state information of the target outlet belonging to the state space; wherein, the state space includes at least one of the following states: historical passenger flow state, predicted passenger flow state for a specified future period, window service state, service provider state, and the state of each IoT device; At least one of the aforementioned state information is input into the decision model to output at least one action information belonging to the action space of the decision model; wherein, the action space includes at least one of the following actions: window service adjustment action, IoT device linkage control action, and service provider guidance action; Based on at least one of the action information, a corresponding action instruction is sent to the controlled object of the action information.

2. The method according to claim 1, characterized in that, The decision-making model of the intelligent agent is obtained in the following way: Construct the state space and action space of the decision model; The state information from the training samples is input into the decision model to output action information determined from the action space; The action information is evaluated based on the reward function, and the decision model is trained based on the evaluation results.

3. The method according to claim 2, characterized in that, The reward function is constructed by weighted summation of multidimensional indicators, including at least one of the following: reduction in waiting time, customer satisfaction, employee workload balancing, and resource switching costs.

4. The method according to claim 3, characterized in that, After evaluating the action information according to the reward function, and before training the decision model based on the evaluation results, the method further includes: Compare the reward with a preset reward distribution range; Rewards that exceed the preset reward distribution range will be filtered out based on the comparison results; After training the decision model based on the evaluation results, the process also includes: The decision-making strategy of the decision-making model and the benchmark rule are compared offline. Based on the offline comparison results, the weights of each indicator in the reward function are manually adjusted to ensure that the decision-making strategy aligns with business priorities.

5. The method according to claim 1, characterized in that, The window service status includes at least one of the following: the type of service currently being processed, the estimated remaining processing time, and the window's availability status; The service provider status includes at least one of the following: service provider location, service provider skills, and service provider idle status; The status of each IoT device includes at least one of the following: idle status of the form filler, occupied status of the seat, and privacy status of the safety glass.

6. The method according to claim 1, characterized in that, Based on at least one of the action information items, send a corresponding action instruction to the controlled object of the action information, including: Based on the window service adjustment action, send a window service adjustment instruction to the queuing system; Based on the IoT device linkage control actions, send device control commands to the IoT device collaborative control network; Based on the actions guided by the service provider, a service instruction is sent to the smart wearable device worn by the target service provider.

7. The method according to claim 6, characterized in that, The state space also includes user characteristics of users within the network; Obtain at least one state information item belonging to the state space of the target network point, including: Get the user characteristics of the current user entering the current branch, the current real-time customer flow distribution, the current status of each IoT device, the current status of each window service, and the current status of each service provider; The decision model is input based on at least one of the aforementioned state information to output at least one action information belonging to the action space of the decision model, including: The user characteristics, the current real-time passenger flow distribution, the current status of each IoT device, the current status of each window service, and the current status of each service provider are input into the decision model to generate window service adjustment instructions, device control instructions, and service instructions. Based on at least one of the action information items, send a corresponding action instruction to the controlled object of the action information, including: Based on the window service adjustment action, send a window service adjustment instruction to the queuing system; Based on the IoT device linkage control actions, send device control commands to the IoT device collaborative control network; Based on the actions guided by the service provider, a service instruction for the current user is sent to the smart wearable device worn by the target service provider.

8. The method according to claim 7, characterized in that, Obtain the user characteristics of the current user entering the current branch, including: The system receives the identification information of the current user reported by the staff at the branch entrance and calls the multimodal perception system to obtain the multimodal biometrics of the current user; the multimodal biometrics include the current user's gait characteristics and / or voice characteristics. The cloud platform is invoked based on the identification information to obtain the current user's historical business information; By comprehensively analyzing the multimodal biometrics of the current user and the historical business information, the user characteristics of the current user are obtained. The user characteristics include the identification information, historical preferences, potential business needs, real-time status index, and business priority. The business priority is determined by a comprehensive evaluation of multiple factors, including business value contribution, criticality and urgency of business type, business complexity, user status, historical risks, and global resource load balancing of the network.

9. A smart service device for service outlets, characterized in that, include: The network information acquisition module is used to acquire at least one state information of the target network belonging to the state space based on the state space of the decision model in the intelligent agent; wherein, the state space includes at least one of the following states: historical passenger flow status, predicted passenger flow status for a specified future period, window service status, service provider status, and status of each Internet of Things device; An action output module is used to input at least one of the aforementioned state information into the decision model, so as to output at least one action information belonging to the action space of the decision model; wherein, the action space includes at least one of the following actions: window service adjustment action, IoT device linkage control action, and service provider guidance action; An action instruction sending module is used to send a corresponding action instruction to the controlled object of the action information based on at least one of the action information.

10. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform a smart service method for service outlets according to any one of claims 1-8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that are used to cause a processor to execute and implement the branch intelligent service method according to any one of claims 1-8.

Images