Adaptive intelligent controlled article integrated management platform and method based on rule engine

Abstract

The application discloses a rule engine-based adaptive intelligent controlled article integrated management platform and method, and relates to the technical field of intelligent supervision of controlled articles. The platform comprises the following modules: multi-source data acquisition, data fusion and context construction, rule engine, risk assessment, business process arrangement, execution linkage, execution result verification, audit traceability and feedback optimization. Through the acquisition of multi-source data such as personnel, articles, inventory, location, equipment, environment and business tasks, the context fact set of operation events is constructed, the rule engine is used to output a basic control strategy, and a target control strategy is generated in combination with risk scoring, so as to drive the execution of operations such as directional opening, limited release and abnormal locking by intelligent cabinets and other equipment, while the actual execution result is verified, a whole-process audit chain is formed, and the rules and thresholds are optimized based on historical results. The application can improve the safety, real-time performance, accuracy and adaptive capacity of controlled article management.

Description

Technical Field

[0001] This invention relates to the fields of intelligent Internet of Things management, data fusion processing, rule engine decision-making, intelligent cabinet control linkage, risk assessment and full-process supervision of controlled items, and particularly to an adaptive intelligent controlled item integrated management platform and method based on a rule engine. Background Technology

[0002] In existing technologies, controlled items are typically managed through manual registration, static permission allocation, fixed approval processes, and post-event inventory and traceability. This management model can meet basic requirements in small-scale, low-frequency, and low-risk scenarios, but it presents the following problems in scenarios involving multiple departments, multiple regions, and multiple types of items circulating in parallel, where security, timeliness, and compliance requirements are high.

[0003] First, the rules are rigid. Existing management systems often write approval rules, requisition rules, return rules, and early warning rules into the program in a fixed business logic manner. This approach makes it difficult to dynamically adjust control strategies according to different departments, time periods, risk levels, inventory statuses, and task urgency, resulting in poor system adaptability.

[0004] Second, the data dimensions are limited. Many existing solutions rely solely on barcodes, RFID, or manually entered data for inbound and outbound registration, lacking the fusion and analysis of multi-dimensional data such as personnel identity, location, time period, task origin, environmental status, equipment status, and historical behavior. Therefore, they cannot accurately identify abnormal behaviors in complex scenarios.

[0005] Third, security and efficiency are difficult to balance. For high-risk controlled items, regulatory requirements often include strict control measures such as dual authentication, multi-person approval, and full traceability; however, in special scenarios such as emergency rescue, disaster relief, on-duty work, nighttime operations, or emergency scientific research, rapid requisition and rapid response are required. The fixed approval path in existing technologies is prone to delays, while simply relaxing permissions can easily lead to compliance risks.

[0006] Fourth, the ability to ensure consistency between accounts and physical inventory is insufficient. Existing technologies typically rely on manual inventory checks, post-event reconciliation, and manual spot checks to detect anomalies. This makes it difficult to identify problems such as over-receiving, mis-receiving or mis-issuing, unauthorized operations, overdue returns, and increased frequency of anomalies in a timely manner, resulting in delayed anomaly detection and high correction costs.

[0007] Fifth, there is a lack of continuous optimization capabilities. The control requirements of different institutions and different management objects vary greatly. The rules for managing controlled items need to be adjusted as business volume changes, anomaly types change, and personnel operating habits change. However, traditional solutions lack the ability to optimize rules based on historical operational data.

[0008] Therefore, how to build a comprehensive management platform and method for controlled items that can achieve multi-source perception, dynamic decision-making, risk classification, equipment linkage, and full-process auditing, and can adaptively adjust control strategies according to real-time scene changes, has become a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0009] To address the shortcomings of existing technologies, this invention provides an adaptive intelligent controlled item integrated management platform and method based on a rule engine, which solves the technical problems existing in the prior art, such as rule rigidity, untimely anomaly identification, uncontrollable entity execution, inconsistency between accounts and actual items, difficulty in balancing approval efficiency and security supervision, and lack of closed-loop optimization.

[0010] This invention achieves adaptive intelligent management of controlled items throughout the entire process of warehousing, storage, application, approval, retrieval, return, destruction, inventory, and auditing by constructing a multi-source data acquisition and fusion mechanism, a contextual fact set construction mechanism, a rule engine decision-making mechanism, a risk assessment and hierarchical control mechanism, an intelligent cabinet control execution linkage mechanism, an execution result verification mechanism, a full-process audit traceability mechanism, and a feedback optimization mechanism.

[0011] To achieve the above objectives, the present invention provides the following technical solution: an adaptive intelligent controlled item integrated management platform based on a rule engine, comprising a multi-source data acquisition module, a data fusion and context construction module, a rule engine module, a risk assessment module, a business process orchestration module, an execution linkage module, an execution result verification module, an audit traceability module, and a feedback optimization module.

[0012] The multi-source data acquisition module is used to collect personnel identity data, item identification data, inventory status data, time data, spatial location data, equipment status data, environmental parameter data, business task data, and historical operation data related to the management of controlled items.

[0013] The data fusion and context construction module is used to standardize, time-align, object-map, identity-bind, and trust-verify the collected multi-source data, and construct a context fact set corresponding to the current controlled item operation event.

[0014] The rule engine module is used to perform rule loading, condition matching, constraint judgment, conflict resolution, and strategy output based on the context fact set to obtain the basic control strategy for the current operation event.

[0015] The risk assessment module is used to score the risk of the current operation event based on the risk factors in the context fact set, generate a risk level, and modify the basic control strategy based on the risk level to obtain the target control strategy.

[0016] The business process orchestration module is used to generate corresponding approval processes, authentication processes, release processes, review processes, return processes, inventory processes, and exception handling processes based on the target control strategy.

[0017] The execution linkage module is used to send process instructions to the controlled item storage and retrieval equipment and external business systems to control at least one of the following operations: opening the target cabinet door, opening the target drawer, releasing a limited quantity, freezing inventory, timeout reminder, and abnormal locking.

[0018] The execution result verification module is used to receive the actual execution data fed back by the controlled item storage and retrieval device and the sensing device, and to compare the actual execution data with the target control strategy to generate a verification result.

[0019] The audit traceability module is used to generate a full-process audit chain record based on the contextual fact set, target control strategy, actual execution data and verification results.

[0020] The feedback optimization module is used to update at least one of the following based on historical audit chain records, rule triggering results, abnormal event statistics, and manual review results: rule parameters, risk thresholds, and process templates.

[0021] Furthermore, the contextual fact set includes at least personnel fact objects, item fact objects, scene fact objects, and event fact objects.

[0022] Furthermore, each rule in the rule engine includes a set of conditions, a set of constraints, a set of actions, exception conditions, and a priority field.

[0023] Furthermore, when multiple rules are triggered simultaneously and the corresponding actions conflict, the rule engine resolves the conflict based on at least two of the following: regulatory compliance level, risk level, scenario urgency, control area level, personnel credit level, and operation sequence.

[0024] Furthermore, the risk assessment module performs a weighted calculation based on at least three risk factors, including the sensitivity level of the controlled item, the current inventory as a percentage of the safety stock, the applicant's authority level, recent operation frequency, cross-regional requisition identifier, abnormal period application identifier, approval completeness, equipment anomaly identifier, environmental anomaly identifier, and the clustering degree of similar anomalies, to generate a risk score and classify the current operation event into one of the following risk levels: low risk, medium risk, high risk, and extremely high risk.

[0025] Furthermore, the execution linkage module is connected to at least one of the following: smart cabinet, smart medicine cabinet, hazardous chemical cabinet, sample cabinet, or lockable storage unit. After receiving the target control strategy, it controls only the target cabinet door or target drawer corresponding to the current operation event to open, while limiting the number of controlled items that can be taken in combination with the authorized quantity.

[0026] Furthermore, the execution result verification module is connected to at least one of the following: RFID identification device, weight sensor, door magnetic sensor, video acquisition device, and environmental sensor. It is used to verify at least one of the following: actual quantity taken out, actual quantity returned, cabinet door opening status, operation time, environmental status, and personnel operation trajectory. When the verification fails, it triggers at least one of the following operations: abnormal locking, audible and visual alarm, message push, supplementary review, and secondary inventory.

[0027] Furthermore, the audit traceability module generates a unique audit chain identifier for each controlled item operation event, and associates at least three of the following records—application record, identity authentication record, approval record, unlocking record, retrieval record, sensor verification record, inventory change record, return record, anomaly handling record, and review record—under the same audit chain identifier.

[0028] Furthermore, the feedback optimization module is used to statistically analyze at least two of the following indicators: rule trigger frequency, false alarm rate, false alarm rate, manual rejection rate, abnormal correlation degree, and process time. Based on these indicators, it updates at least one of the following: rule priority, rule threshold, approval level, and exception conditions, or generates rule optimization suggestions for management confirmation.

[0029] This invention also provides an adaptive intelligent controlled item integrated management method based on a rule engine, applied to the aforementioned platform, comprising the following steps: Collect multi-source data related to the operation events of the currently controlled items; The multi-source data is standardized, time-aligned, object-mapped, identity-bound, and trusted for verification to construct a contextual fact set of the current controlled item operation event; The context fact set is input into the rule engine to perform rule matching, constraint judgment, conflict resolution and output basic control strategy; Risk scores are generated based on risk factors in the contextual fact set, and risk levels are then generated. The basic control strategy is then modified based on the risk levels to generate the target control strategy. Based on the target control strategy, business processes are arranged and execution instructions are issued to control the storage and retrieval equipment of controlled items and the external business systems to perform corresponding operations; Collect actual execution data and verify the consistency between the actual execution data and the target control strategy; Update the status of controlled items, inventory ledger, and audit chain records based on the verification results; When verification fails or an abnormal condition is met, the exception handling process is triggered. Based on historical audit chain records and anomaly handling results, at least one of the following should be optimized: rule parameters, risk thresholds, and process templates.

[0030] The present invention also provides an electronic device, including a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the computer program implements the above-described method when executed by the processor.

[0031] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method.

[0032] Compared with existing technologies, this invention provides an adaptive intelligent controlled item integrated management platform and method based on a rule engine, which has the following beneficial effects: By integrating multi-source data and constructing contextual fact sets, the management of controlled items is transformed from single-point registration to dynamic and comprehensive perception of people, items, scenarios, and events, thereby improving the accuracy of decision-making in complex scenarios.

[0033] Through a two-tiered decision-making mechanism of rules engine and risk assessment, the system can not only execute preset rules, but also adjust the control intensity according to the real-time risk level, thereby balancing security supervision and business efficiency.

[0034] Through rule conflict resolution and exception scenario switching mechanisms, the system can generate differentiated control strategies in special situations such as emergency rescue, night shift, temporary authorization, and remote authorization, thereby improving adaptability and practicality.

[0035] By linking the execution module with physical devices such as smart cabinets, a closed loop is achieved from information system decision-making to physical space control, which can directly limit the number of cabinet doors, drawers, and items taken out, thereby improving execution controllability.

[0036] The execution result verification module performs real-time verification of the taking and putting actions, quantities, durations, and environmental status, thereby improving the consistency between accounts and actual items and reducing the risks of unauthorized taking, incorrect taking and putting, and missing records.

[0037] By recording the entire audit chain, the entire process of application, authentication, approval, unlocking, retrieval, return, and exception handling can be linked together, significantly enhancing traceability and audit credibility.

[0038] Through the feedback optimization module, the platform can continuously optimize rule parameters, process templates and risk thresholds based on historical operating results, so that the system can maintain good adaptability and control effect after long-term operation. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of the overall structure of the adaptive intelligent controlled item integrated management platform based on a rule engine in an embodiment of the present invention.

[0040] Figure 2This is a flowchart illustrating the comprehensive management method for controlled items in an embodiment of the present invention.

[0041] Figure 3 This is a schematic diagram illustrating the context fact set construction process in an embodiment of the present invention.

[0042] Figure 4 This is a schematic diagram of the two-layer decision-making process of rule engine and risk assessment in an embodiment of the present invention.

[0043] Figure 5 This is a schematic diagram of the closed loop of execution linkage and execution result verification in an embodiment of the present invention.

[0044] Figure 6 This is a schematic diagram of the audit traceability chain structure in an embodiment of the present invention.

[0045] Figure 7 This is a schematic diagram of the feedback optimization mechanism in an embodiment of the present invention. Detailed Implementation

[0046] To better understand the purpose, structure, and function of this invention, the following detailed description of the rule engine-based adaptive intelligent controlled item integrated management platform and method is provided in conjunction with the accompanying drawings.

[0047] Please see Figure 1-7 This invention: Embodiment 1: Platform Overall Architecture like Figure 1 As shown, this embodiment provides an adaptive intelligent controlled item integrated management platform based on a rule engine, including a perception access layer, a data fusion layer, a decision control layer, a business execution layer, and a supervision and optimization layer.

[0048] The perception access layer is used to connect various front-end devices and external systems, including but not limited to RFID readers, barcode or QR code scanning devices, smart cabinet lock control units, weight sensors, door magnetic sensors, video acquisition devices, face recognition terminals, fingerprint or finger vein authentication terminals, environmental sensors, mobile terminals, as well as hospital HIS systems, laboratory management systems, warehousing systems, ERP systems, approval systems, etc.

[0049] The data fusion layer is used for unified access and preprocessing of data from different sources, completing encoding conversion, protocol adaptation, timestamp alignment, object identification, identity binding, and trust verification. This layer uniformly maps the collected data into personnel objects, item objects, scene objects, event objects, and risk objects.

[0050] The decision control layer includes a rules engine module and a risk assessment module. The rules engine module outputs basic control strategies based on the contextual fact set, while the risk assessment module modifies the basic control strategies to determine the final target control strategy.

[0051] The business execution layer includes a business process orchestration module, an execution linkage module, and an execution result verification module, which are used to drive operations such as approval, authentication, cabinet opening, release, reminders, locking, and review.

[0052] The supervision and optimization layer includes an audit traceability module and a feedback optimization module, which are used to form a full-chain audit record and continuously optimize rules, thresholds and processes.

[0053] In this embodiment, the platform can be deployed in a local private network environment or in a cloud-edge collaborative deployment mode. For cabinet control commands, door magnetic detection, and timeout locking actions with high real-time requirements, it is preferred to execute them on the edge control node; for rule statistics, model analysis, and strategy optimization, it is preferred to execute them on the central server.

[0054] Example 2: Context Fact Set Construction Mechanism like Figure 3 As shown, after the platform receives an operation request for a controlled item, the multi-source data acquisition module first collects relevant data, and then the data fusion and context construction module constructs a context fact set for the current operation event.

[0055] In a preferred embodiment, the contextual fact set includes at least the following four types of fact objects.

[0056] Personnel Facts This includes information such as the user's unique identifier, department, job role, authorization level, training qualifications, on-duty status, historical operation frequency, historical violation count, credit rating, and most recent abnormal correlation.

[0057] Item fact object This includes information such as the item's unique identifier, item name, category, specifications, batch number, expiration date, sensitivity level, current inventory, safety stock, storage cabinet number, target drawer number, and lifecycle status.

[0058] Scene Fact Object Information includes the current time period, geographical region, operating terminal, environmental parameters, equipment status, task source, business urgency, whether it belongs to an emergency scenario, whether it belongs to cross-regional operation, and whether it belongs to nighttime operation.

[0059] Event Fact Object It includes at least one of the following actions: application, authentication, approval, unlocking, retrieval, return, destruction, and inventory, and records the event start time, target quantity, associated document number, and associated task number.

[0060] In a specific example, when the system receives a request for controlled substances from an emergency room doctor, it simultaneously retrieves the doctor's facial recognition results, employee ID login record, shift task information, corresponding medical orders, current medicine cabinet inventory information, medicine cabinet door sensor status, current time period, emergency room area status, and statistics on the doctor's frequent medication retrieval behavior over the past thirty days. This data is then used to construct a complete set of contextual facts. This set of facts serves as the foundation for subsequent rule-based reasoning and risk scoring.

[0061] In this way, the platform no longer relies solely on data from a single application form, but instead establishes a structured factual basis that is highly relevant to the current operational scenario, thereby improving the quality of subsequent decision-making.

[0062] Example 3: Rule Engine Mechanism like Figure 4 As shown, the rule engine module in this embodiment includes a rule modeling unit, a fact loading unit, a matching reasoning unit, a conflict resolution unit, and a policy output unit.

[0063] Each rule should preferably adopt the following structure: The condition set is used to define the basic conditions for triggering a rule; Constraint sets are used to define the restrictions that must be met when a rule is executed. Action set, used to define the control actions output after a rule is hit; Exception conditions are used to define alternative processing strategies for special scenarios; The priority field is used to define the rule ordering and conflict resolution weights.

[0064] For example, for scenarios involving the requisition of ordinary high-risk items, the following rule logic can be configured: When a person's role is an authorized position, the business task has a legitimate source, the inventory is higher than the safety stock, the current area is in a normal state, and the current time is within the allowed time period, if the person has no overdue return records, does not exceed the daily frequency threshold, and has passed two-factor authentication, then the action set of "supervisor approval, target drawer opening, limited release, video recording, return timer start" will be output.

[0065] For emergency rescue scenarios, exception rules can be configured: when an event is marked as an emergency rescue task and the applicant is on the emergency rescue authorization whitelist, it is allowed to add post-event signature, video recording, and mandatory review after timeout while reducing the number of approval levels.

[0066] When multiple rules are triggered simultaneously, such as the "emergency priority rule" and the "high-risk drug dual signature rule" being triggered at the same time, the conflict resolution unit sorts and calculates based on the regulatory compliance level, risk level, scenario urgency, and personnel credit level, and generates a compromise control strategy, such as "first release by designated personnel, then complete remote dual signature, and simultaneously trigger full video recording and time-limited review."

[0067] Through the above design, the rule engine is no longer a simple static approval logic, but can provide context-driven adaptive output for different scenarios.

[0068] Example 4: Risk Assessment and Tiered Control Mechanism To further avoid the rigid control problems caused by relying solely on rule conditions, this embodiment sets up a risk assessment module outside of the rule engine.

[0069] The risk assessment module extracts risk factors from the contextual fact set, including but not limited to the following: Item sensitivity level; The ratio of current inventory to safety stock; Does the applicant's access level match the sensitivity level of the items? Has the recent trading frequency been abnormal? Is it a cross-regional pickup? Is it during an unusually long period? Is the current approval chain complete? Is the equipment in an abnormal state? Are the environmental conditions exceeding the limits? Have similar abnormal events occurred in clusters recently?

[0070] In a preferred embodiment, the risk score can be obtained by a weighted summation, as shown in the following example: R = a1×F1 + a2×F2 + a3×F3 + … + an×Fn Where R represents the risk score, F1 to Fn represent the quantified values ​​of different risk factors, and a1 to an represent the corresponding weight parameters. The weights can be preset by the administrator or fine-tuned within the authorized range by the feedback optimization module.

[0071] Based on the risk scoring results, events are classified into four levels: low risk, medium risk, high risk, and extremely high risk, and the following controls are implemented: Low risk: Automatic approval, recording routine audit information; Medium risk: Requires secondary certification or supervisor approval; High-risk: Double verification, limited distribution, and full process record keeping; Extremely high risk: Suspension of issuance, manual verification, and incident reporting.

[0072] In a specific example, if a hazardous chemical is requested at night by personnel from outside the laboratory, and the inventory is close to the safety stock limit, and environmental sensors detect an abnormal temperature in the target cabinet area, then even if the system hits some release rules, the basic control strategy will be upgraded to "suspension of release + manual verification" due to the increased risk score.

[0073] By using a rule engine to output basic strategies and a risk assessment module to correct them, this invention achieves a two-tier decision-making structure of "rule-driven + risk correction".

[0074] Example 5: Execution Linkage and Entity Control Mechanism like Figure 5 As shown, after the target control strategy is formed, the business process orchestration module parses it into a series of executable operations, which are then distributed to smart devices and external systems by the execution linkage module.

[0075] In a preferred embodiment, the controlled item storage and retrieval device is a smart cabinet or a smart medicine cabinet. Each cabinet includes a door locking unit, a drawer locking unit, an RFID identification unit, a weight detection unit, a door magnetic detection unit, and a local controller.

[0076] Once the system receives an approved and risk-corrected target control policy, it will only open the target cabinet door or drawer related to the current operational event, and will not allow unrelated cabinets to be opened simultaneously. If the policy includes quantity restrictions, the system will send authorized quantity parameters to the local controller before the cabinet door is opened; the local controller will then monitor the actual quantity taken out in real time, combining weight change values ​​or RFID reading results.

[0077] For example, when 3 controlled reagents are authorized to be taken out, if the system detects that 5 reagents have actually been taken out, the local controller can immediately perform at least one of the following actions: lock the drawer, provide a voice prompt, issue an audible and visual alarm, report the abnormal event, or request verification and confirmation.

[0078] Upon return, the system will only open the corresponding return slot or designated drawer, and will verify whether the return behavior is consistent with the original requisition record by reading batch information, weight detection quantity information, and time window information through RFID. If it is found that the returned items are inconsistent with the original batch, the returned quantity is insufficient, or the return time limit is exceeded, an abnormal process will be automatically triggered.

[0079] In this way, the present invention transforms the rule-based decision results into fine-grained control at the physical space level, thereby improving the rigidity and accuracy of execution.

[0080] Example 6: Execution Result Verification and Anomaly Closure The execution result verification module is used to receive feedback data from the smart cabinet and sensing devices in real time, including the number of items taken out and put in, cabinet door status, door opening time, identification tags, environmental parameters, and video event indexes.

[0081] The system compares the feedback data with the target control strategy to generate a verification result. If the verification result is consistent, the inventory ledger, item lifecycle status, and audit chain records are updated. If the verification result is inconsistent, different closed-loop processing is performed based on the anomaly type.

[0082] Exception types may include, but are not limited to: The amount withdrawn exceeded the authorized limit; The retrieved item does not match the authorized item; The cabinet door was left open for an extended period without being closed. The return has not been completed and the stipulated time limit has been exceeded; Return batch mismatch; The personnel's trajectory does not match their authorized location; Abnormal environmental parameters; The equipment is malfunctioning; The approval process is missing.

[0083] For different anomalies, the system may perform one or more of the following actions in combination: Lock cabinet doors or drawers; Issue an audible and visual alarm; Push messages to supervisors or safety officers; Initiate two-factor authentication; Generate supplementary review tasks; Triggering a second inventory count; Suspend subsequent similar requisitions; The incident was escalated and reported.

[0084] In one implementation example, if hazardous chemicals in the laboratory are not returned within the specified time and environmental sensors detect abnormal temperature changes in the relevant area, the system will raise the risk level of the event and simultaneously push a message to the laboratory manager, safety officer, and equipment administrator, requiring on-site verification.

[0085] By implementing a result verification and anomaly closed-loop processing mechanism, this invention enables rapid detection, rapid blocking, and rapid handling of anomalies.

[0086] Example 7: Audit Traceability Mechanism like Figure 6 As shown, the audit traceability module generates a unique audit chain identifier for each controlled item operation event and links the key records related to the event to this identifier.

[0087] The key records include, but are not limited to: Application records; Identity authentication records; Approval records; Cabinet door opening record; Drawer operation log; RFID reading and recording; Record of weight changes; Door sensor status record; Video index records; Inventory change records; Return records; Destroy records; Exception handling log; Review record.

[0088] In a preferred embodiment, the audit traceability module adds timestamps and summary verification information to key records and strings them together according to event sequence to form an inseparable operational chain. For high-risk operations, an anti-tampering log storage mechanism can also be used to enhance record credibility.

[0089] This mechanism can solve problems in traditional systems such as "disconnect between approval records and actual use", "records in the ledger but lack of on-site evidence", and "inability to restore the complete process after an anomaly is discovered".

[0090] Example 8: Feedback Optimization Mechanism like Figure 7 As shown, the feedback optimization module continuously analyzes the system's performance based on historical audit chain records, abnormal event statistics, manual review results, and process operation indicators.

[0091] The statistical indicators may include: Hit frequency of each rule; Rule false alarm rate; Rule-based false negative rate; Manual rejection rate; Correlation degree of a certain type of abnormal events; Average approval time; Blockage rate of a certain type of process; Equipment linkage failure rate; Trends in the consistency rate between accounts and actual inventory.

[0092] The feedback optimization module can perform at least one of the following operations based on the above indicators: Adjust rule priority; Adjust the risk scoring threshold; Adjust the control strength for exceptional scenarios; Adjust the approval level; Optimization suggestions for generation rules are provided for administrator confirmation; Automatic fine-tuning of low-risk parameters within the authorized range.

[0093] For example, if the system consistently finds that a certain type of low-risk item shows almost no anomalies during night shifts, but the response is slow due to too many approval levels, the feedback optimization module can suggest changing the approval path in night shifts from dual approval to single approval. As another example, if a certain type of hazardous chemical frequently experiences discrepancies between records and actual stock during a certain period, the module can suggest increasing the priority of the corresponding rule and increasing the frequency of random reviews during that period.

[0094] Through the above mechanism, the present invention possesses adaptive optimization capabilities during long-term operation.

[0095] Example 9: Medical Psychotropic Drug Management Scenario This embodiment illustrates the application of the present invention in a hospital setting.

[0096] An emergency room doctor initiates a request to requisition a certain controlled substance during a night shift. The system first obtains the person's identity information through a facial recognition terminal, and retrieves their employee ID, duty status, and department information; simultaneously, it retrieves the corresponding medical order, patient information, and emergency task marker from the HIS system; it obtains the inventory status, target drawer status, and door sensor status from the medicine cabinet; and it retrieves statistics on the doctor's recent high-frequency medication requisition behavior from the log system.

[0097] The data fusion and context building module generates a contextual fact set based on this. The rule engine determines that the event belongs to an emergency scenario, triggering the emergency exception rule; the risk assessment module further determines that although it is a highly sensitive drug, the applicant's qualifications are complete, the source of the task is clear, the inventory is sufficient, and the equipment is in normal condition. Therefore, the risk level is determined to be between medium and high, and the basic control strategy is revised to "rapid certification + targeted cabinet opening + strong video record + post-event signing + time-limited review".

[0098] The system activates a linked module to open a designated drawer, allowing only the authorized quantity of medication to be retrieved. Weight sensors and RFID identification units verify the quantity and batch number of medications retrieved in real time. If the retrieval is complete and the verification is successful, the inventory is automatically deducted, an audit chain is automatically generated, and a timer for recording the return or usage result is initiated. If the result is not recorded within the time limit, the system automatically escalates the risk level and sends a review task to the pharmacy department and department head.

[0099] This embodiment demonstrates the technical advantages of the present invention in scenarios where high timeliness and high compliance coexist.

[0100] Example 10: Laboratory Hazardous Chemicals Management Scenario This embodiment illustrates the application of the present invention in the management of hazardous chemicals in laboratories.

[0101] Researchers apply for a hazardous chemical for a designated experimental project. The system retrieves the applicant's training qualifications, affiliated laboratory, project number, intended use, expected usage, and location, and constructs a contextual fact set by combining this information with the chemical's sensitivity level, current inventory, safety stock, target enclosure environmental parameters, and historical anomaly records.

[0102] The rules engine determines that the applicant has basic permissions, but the constraints require that the training record be valid, the project approval be valid, and the laboratory environment be normal. If any of these conditions are not met, the rules engine outputs a restrictive basic policy. The risk assessment module comprehensively considers that the application involves cross-regional use, the current time is night, and the target cabinet temperature is slightly higher than the normal threshold, raising the risk level to high risk, and thus modifying the target control policy to "suspend automatic release + safety officer review + targeted cabinet opening after on-site confirmation".

[0103] If the environmental parameters subsequently return to normal and pass the review, the system will then implement limited release. Upon return, the container tag is identified via RFID, and the batch and remaining quantity are verified. If there is a discrepancy, the system will enter an abnormal storage phase and trigger manual verification.

[0104] This embodiment demonstrates the invention's ability to precisely control high-risk laboratory items.

[0105] Example 11: Controlled Classified Media Management Scenario This embodiment illustrates the application of the present invention in the management of classified carriers.

[0106] Classified documents are stored in intelligent secure cabinets, and users must initiate a retrieval request through an authorized terminal. The system collects information such as the applicant's security classification, job authority, current location, reason for request, corresponding task sheet, secure cabinet status, and video surveillance status to construct a contextual fact set.

[0107] The rules engine outputs basic policies based on the level of confidentiality and the urgency of the task. The risk assessment module further combines cross-regional data retrieval, abnormal time periods, and recent operation frequency to form a risk judgment. If the risk is high, the system requires dual authentication and supervisor approval; if it is a registered urgent task, it allows for rapid targeted opening of the cabinet according to the exception rules, but requires simultaneous video recording and completion of return or transfer registration within a limited time.

[0108] As can be seen from this embodiment, the present invention is not only applicable to medical and laboratory scenarios, but also to other types of items that require controlled management.

[0109] It is understood that the present invention has been described through some embodiments, and those skilled in the art will recognize that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the invention. Furthermore, under the teachings of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the protection scope of the present invention.

Claims

1. An adaptive intelligent controlled item integrated management platform based on a rule engine, characterized in that: include: The multi-source data acquisition module is used to collect personnel identification data, item identification data, inventory status data, time data, spatial location data, equipment status data, environmental parameter data, business task data, and historical operation data related to the management of controlled items. The data fusion and context building module is used to perform standardization, time alignment, object mapping, identity binding and trust verification on the data collected by the multi-source data acquisition module, and generate a context fact set corresponding to the current controlled item operation event; The rules engine module is used to perform rule loading, condition matching, constraint judgment, conflict resolution, and strategy output based on the context fact set to obtain the basic control strategy for the operation event of the currently controlled item; The risk assessment module is used to score the risk of the current controlled item operation event based on the risk factors in the context fact set, generate a risk level, and modify the basic control strategy based on the risk level to obtain the target control strategy. The business process orchestration module is used to generate corresponding approval processes, authentication processes, release processes, review processes, return processes, inventory processes, and exception handling processes based on the target control strategy. The execution linkage module is used to send the process instructions output by the business process orchestration module to the controlled item storage and retrieval equipment and / or external business system, so as to control the controlled item storage and retrieval equipment to perform at least one of the following operations: target cabinet door opening, target drawer opening, limited quantity release, inventory freezing, timeout reminder and abnormal locking. The execution result verification module is used to receive actual execution data fed back by the controlled item storage and / or sensing device, and compare the actual execution data with the target control strategy to generate a verification result; The audit traceability module is used to generate a full-process audit chain record based on the context fact set, the target control strategy, the actual execution data, and the verification results. The feedback optimization module is used to update at least one of the following based on historical audit chain records, rule trigger results, abnormal event statistics, and manual review results: rule parameters, risk thresholds, and process templates.

2. The adaptive intelligent controlled item integrated management platform based on a rule engine according to claim 1, characterized in that, The context fact set constructed by the data fusion and context construction module includes at least two of the following fact objects: Personnel fact objects are used to characterize the operator's identity, role, position, qualifications, on-duty status, historical credit rating, and historical violation records; Item fact objects are used to characterize the category, batch, specifications, sensitivity level, expiration date, current inventory, storage location, and lifecycle status of controlled items; Scene fact objects are used to represent the time interval, spatial region, environmental state, device state, task source, and business urgency corresponding to the current operation; An event fact object is used to represent at least one of the following actions in the current operation: application action, approval action, unlocking action, retrieval action, return action, destruction action, and inventory action.

3. The adaptive intelligent controlled item integrated management platform based on a rule engine according to claim 1, characterized in that, Each rule in the rule engine module includes at least a set of conditions, a set of constraints, a set of actions, exception conditions, and a priority field. The condition set is used to represent the basic scenario conditions for rule triggering, the constraint set is used to represent the restriction conditions that must be met before the rule is executed, the action set is used to represent the control action that needs to be output after the rule is hit, the exception condition is used to represent at least one special scenario among emergency scenario, night scenario, temporary authorization scenario and remote authorization scenario, and the priority field is used to determine the rule execution order when multiple rules are hit at the same time.

4. The adaptive intelligent controlled item integrated management platform based on a rule engine according to claim 3, characterized in that, The rule engine module also includes a conflict resolution unit. When multiple rules are hit simultaneously and the corresponding actions conflict, the conflict resolution unit performs sorting calculations based on at least two of the following: regulatory compliance level, risk level, scenario urgency, control area level, personnel credit level, and operation sequence, in order to output the basic control strategy after conflict resolution.

5. The adaptive intelligent controlled item integrated management platform based on a rule engine according to claim 1, characterized in that, The risk assessment module performs a weighted calculation based on at least three risk factors, including the sensitivity level of the controlled item, the current inventory as a percentage of the safety stock, the applicant's authority level, recent operation frequency, cross-regional requisition identifier, abnormal period application identifier, approval completeness, equipment anomaly identifier, environmental anomaly identifier, and the clustering degree of similar anomalies, to generate a risk score. Based on the risk score, the current controlled item operation event is classified into one of the following risk levels: low risk, medium risk, high risk, and extremely high risk.

6. The adaptive intelligent controlled item integrated management platform based on a rule engine according to claim 1, characterized in that, The execution linkage module is connected to the controlled item storage and retrieval device, which includes at least one of the following: intelligent cabinet, intelligent medicine cabinet, hazardous chemical cabinet, sample cabinet, or lockable storage unit. Upon receiving the target control strategy, the execution linkage module only controls the opening of the target cabinet door or target drawer corresponding to the operation event of the currently controlled item, and limits the number of items that can be taken in combination with the authorized quantity in the target control strategy.

7. The adaptive intelligent controlled item integrated management platform based on a rule engine according to claim 6, characterized in that, The execution result verification module is connected to at least one of the following: RFID identification device, weight sensor, door magnetic sensor, video acquisition device, and environmental sensor. The execution result verification module is used to verify at least one of the following based on the actual execution data fed back by the at least one device: the actual number of controlled items taken out, the actual number returned, the cabinet door opening status, the operation time, the environmental status, and the personnel operation trajectory. When the verification fails, the execution linkage module is triggered to perform at least one of the following operations: abnormal locking, audible and visual alarm, message push, supplementary review, and secondary inventory.

8. The adaptive intelligent controlled item integrated management platform based on a rule engine according to claim 1, characterized in that, The audit traceability module generates a unique audit chain identifier for each controlled item operation event and associates at least three of the following records—application record, identity authentication record, approval record, unlocking record, retrieval record, sensor verification record, inventory change record, return record, anomaly handling record, and review record—under the same audit chain identifier to form a traceable, full-process audit chain record.

9. The adaptive intelligent controlled item integrated management platform based on a rule engine according to claim 1, characterized in that, The feedback optimization module is used to calculate at least two of the following metrics for each rule: trigger frequency, false alarm rate, false negative rate, manual rejection rate, anomaly correlation, and process time. Based on the aforementioned metrics, at least one of the following in the rule engine module—rule priority, rule threshold, approval level, and exception conditions—is updated, or rule optimization suggestions are generated for management confirmation.

10. A comprehensive adaptive intelligent controlled item management method based on a rule engine, characterized in that, Applied to the platform as described in any one of claims 1 to 9, comprising the following steps: Collect multi-source data related to the current controlled item operation events, wherein the multi-source data includes at least personnel identification data, item identification data, and inventory status data; The multi-source data is standardized, time-aligned, object-mapped, identity-bound, and trusted for verification to construct a contextual fact set of the current controlled item operation event; The context fact set is input into the rule engine to perform rule matching, constraint judgment, conflict resolution and output basic control strategy; Risk scoring is performed based on the risk factors in the context fact set to generate a risk level, and the basic control strategy is modified according to the risk level to generate a target control strategy. The business process is arranged according to the target control strategy and execution instructions are issued to control the controlled item storage and retrieval equipment and / or external business systems to perform corresponding operations; Collect actual execution data and perform consistency verification between the actual execution data and the target control strategy; Update the status of controlled items, inventory ledger, and audit chain records based on the verification results; When verification fails or an abnormal condition is met, the exception handling process is triggered. Optimize at least one of the rule parameters, risk thresholds, and process templates based on historical audit chain records and anomaly handling results. The step of modifying the basic control strategy based on the risk level to generate the target control strategy includes: When the risk level is low, a control strategy is generated that automatically passes and records routine audit information. When the risk level is medium risk, generate control strategies that require additional certification or supervisor approval. When the risk level is high, a control strategy is generated that includes double review, full traceability, and limited distribution. When the risk level is extremely high, control strategies such as suspending issuance, manual verification, and incident reporting are generated.

Images