Method and device for controlling financial risks of robot leasing business and storage medium

Abstract

The application provides a financial risk control method and device for a robot leasing business and a storage medium, relates to the technical field of data processing, and comprises the following steps: determining an input state of a financial risk control agent by taking an operating health score of a target robot, an equipment evaluation value, a multi-dimensional risk early warning result and a credit score of an enterprise leasing the target robot as input conditions and obtaining a financial risk control decision result for different scenarios output by the financial risk control agent according to the input conditions; wherein the multi-dimensional risk early warning result comprises risk early warning results of at least two dimensions, i.e., an equipment operation dimension, an enterprise operation dimension, an enterprise credit dimension and an equipment value dimension; and the scheme of the application can solve the problem of insufficient comprehensiveness and accuracy of the financial risk control decision result of the robot leasing business.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and specifically to a method, apparatus, and storage medium for financial risk control in robot leasing business. Background Technology

[0002] Because companies need to invest heavily in the purchase and use of robots, they often choose to lease robots to reduce this investment. Leasing robots can alleviate the initial financial pressure on companies and allow for flexible adjustments to the number of leased robots, thus greatly reducing the risk of fixed asset investment.

[0003] However, when robots are currently leased to enterprises, their financial risk control is mostly based on a single data dimension, resulting in insufficient comprehensiveness and accuracy of financial risk control decision-making results. Summary of the Invention

[0004] At least one embodiment of the present invention provides a method, apparatus and storage medium for financial risk control in robot leasing business, which is used to solve the problem that the financial risk control decision results in robot leasing business are not comprehensive and accurate enough in the prior art.

[0005] To solve the above-mentioned technical problems, the present invention is implemented as follows:

[0006] In a first aspect, embodiments of the present invention provide a method for controlling financial risks in robot leasing business, including:

[0007] The operational health score of the target robot, the equipment assessment value, and the multi-dimensional risk warning results, as well as the credit score of the enterprise leasing the target robot, are determined as the input state of the financial risk control intelligent agent, and the financial risk control decision results for different scenarios are obtained by the financial risk control intelligent agent based on the input state.

[0008] The multi-dimensional risk warning results include risk warning results in at least two of the following dimensions: equipment operation dimension, enterprise operation dimension, enterprise credit dimension, and equipment value dimension.

[0009] Optionally, the financial risk control method for the robot leasing business further includes at least two of the following:

[0010] Based on at least one of the following in the target robot's operational status data: equipment utilization rate, capacity utilization rate, failure frequency, maintenance frequency, operating load, and uptime, a risk warning result for the equipment operation dimension is obtained.

[0011] Based on the enterprise's operational data, risk warning results for the enterprise's operational dimensions are obtained;

[0012] A credit rating is obtained based on the credit score, and a risk warning result for the enterprise's credit dimension is obtained based on the credit score and / or credit rating.

[0013] Based on the assessed value of the equipment, risk warning results are obtained for the equipment value dimension.

[0014] Optionally, the financial risk control method for the robot leasing business further includes:

[0015] For at least two of the following dimensions—equipment operation dimension, enterprise operation dimension, enterprise credit dimension, and equipment value dimension—a sliding time window model corresponding to the risk warning conditions is configured. The sliding time window model includes at least one of the following: window length, sliding step size, and data sampling frequency.

[0016] Based on the sliding time window model, acquire the enterprise's business data, the target robot's operating status data, and the robot's macro market data.

[0017] Optionally, the financial risk control method for the robot leasing business further includes:

[0018] The operational health score is obtained based on the target robot's operational status data and a preset operational health assessment model.

[0019] Optionally, the financial risk control method for the robot leasing business further includes:

[0020] The appraised value of the equipment is obtained based on at least one of the following from the macro market data on robots: initial value of the equipment, basic annual depreciation rate, and market conditions; the uptime and failure frequency from the operating status data of the target robot; and a preset equipment value assessment model.

[0021] Optionally, the financial risk control method for the robot leasing business further includes:

[0022] The enterprise's business data and the target robot's operating status data are respectively converted into feature vectors, and the converted feature vectors are weighted and summed to obtain a fused feature vector.

[0023] The credit score is obtained based on the fused feature vector and the preset machine learning model.

[0024] Secondly, embodiments of the present invention also provide a financial risk control device for robot leasing business, comprising:

[0025] The control module is used to determine the operational health score of the target robot, the equipment assessment value, and the multi-dimensional risk warning results, as well as the credit score of the enterprise leasing the target robot, as the input state of the financial risk control intelligent agent, and to obtain the financial risk control decision results for different scenarios output by the financial risk control intelligent agent based on the input state.

[0026] The multi-dimensional risk warning results include risk warning results in at least two of the following dimensions: equipment operation dimension, enterprise operation dimension, enterprise credit dimension, and equipment value dimension.

[0027] Thirdly, embodiments of the present invention also provide a financial risk control device for robot leasing business, comprising: a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the processor executes the program or instructions to implement the financial risk control method for robot leasing business as described in the first aspect.

[0028] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the financial risk control method for robot leasing business as described in the first aspect.

[0029] Fifthly, embodiments of the present invention also provide a computer program product, including computer instructions, which, when executed by a processor, implement the financial risk control method for robot leasing business as described in the first aspect.

[0030] Compared with existing technologies, embodiments of the present invention provide a financial risk control method, device, and storage medium for robot leasing businesses. The method determines the operational health score, equipment valuation, and multi-dimensional risk warning results of the target robot, as well as the credit score of the company leasing the target robot, as the input state of a financial risk control intelligent agent. The agent then obtains financial risk control decision results for different scenarios based on the input state. The multi-dimensional risk warning results include at least two of the following dimensions: equipment operation dimension, company operation dimension, company credit dimension, and equipment value dimension. This approach, by obtaining financial risk control decision results from at least two of these dimensions, solves the problem of insufficient comprehensiveness and accuracy of financial risk control decision results in robot leasing businesses. Attached Figure Description

[0031] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0032] Figure 1 This is a flowchart illustrating the financial risk control method for robot leasing business as described in an embodiment of the present invention.

[0033] Figure 2 This is a flowchart illustrating one embodiment of the financial risk control method for robot leasing business described in this invention.

[0034] Figure 3 This is a schematic diagram of the architecture of the application system for the financial risk control method for robot leasing business described in this embodiment of the invention;

[0035] Figure 4 This is a schematic diagram of the modules of the financial risk control device for robot leasing business according to an embodiment of the present invention;

[0036] Figure 5 This is a hardware block diagram of the financial risk control device for the robot leasing business described in an embodiment of the present invention. Detailed Implementation

[0037] To make the technical problems, technical solutions, and advantages of the present invention clearer, a detailed description will be provided below in conjunction with the accompanying drawings and specific embodiments.

[0038] In various embodiments of the present invention, it should be understood that the sequence number of each process described below does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0039] In addition, the terms "system" and "network" are often used interchangeably in this article.

[0040] The terms "first," "second," etc., used in the specification and claims of this invention are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first" and "second" are generally of the same class, not limited in number; for example, the first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0041] Reference Figure 1 This invention provides a method for controlling financial risks in robot leasing businesses, comprising:

[0042] Step 101: The operational health score of the target robot, the equipment assessment value, and the multi-dimensional risk warning results, as well as the credit score of the enterprise leasing the target robot, are determined as the input state of the financial risk control intelligent agent, and the financial risk control decision results for different scenarios are obtained by the financial risk control intelligent agent based on the input state.

[0043] The multi-dimensional risk warning results include risk warning results in at least two of the following dimensions: equipment operation dimension, enterprise operation dimension, enterprise credit dimension, and equipment value dimension.

[0044] In one embodiment, optionally, before step 101, the method further includes:

[0045] Collect the enterprise's business operation data and connect to data platforms through preset interfaces, including but not limited to: industrial and commercial data platforms, judicial data platforms, tax data platforms, social security data platforms, and water and electricity data platforms. Collect the enterprise's business operation data, including but not limited to: basic enterprise information, litigation records, tax payment status, and operating costs.

[0046] Collect the operating status data of the target robot, deploy an IoT data acquisition terminal, establish a communication connection with the target robot, and collect the real-time operating status data, including but not limited to: power-on time, operating load, fault records, equipment utilization rate, capacity utilization rate, and maintenance history;

[0047] Collect macro market data on robots by using web crawlers or industry database interfaces, including but not limited to: industry prosperity index, regional policy changes, and second-hand equipment market conditions;

[0048] The enterprise operation data, the operation status data, and the robot macro market data are preprocessed, including but not limited to: cleaning, deduplication, format conversion, and missing value filling, in order to remove noise interference and ensure data quality.

[0049] In this embodiment of the invention, the perception layer in the financial risk control system of the robot leasing business can provide multi-dimensional data input to the financial risk control intelligence, enabling the financial risk control intelligence to have comprehensive data perception of enterprise operations, equipment operation, and macro market. All collected and preprocessed data are transmitted to the core data platform of the financial risk control intelligence for unified management and scheduling, realizing standardized data flow access. Each piece of data is timestamped, labeled with indicator dimensions, uniquely identified by the equipment, and uniquely identified by the enterprise, ensuring the uniqueness and traceability of the data.

[0050] In one embodiment, optionally, before step 101, the method further includes:

[0051] The enterprise's business data and the target robot's operating status data are respectively converted into feature vectors, and the converted feature vectors are weighted and summed to obtain a fused feature vector.

[0052] The credit score is obtained based on the fused feature vector and the preset machine learning model.

[0053] In this embodiment of the invention, a corporate credit scoring model can be provided. This model instructs the following steps: first, the company's operational data and the target robot's operational status data are respectively converted into feature vectors; then, the converted feature vectors are weighted and summed to obtain a fused feature vector; finally, the fused feature vector is input into a preset machine learning model; and the machine learning model outputs the credit score based on the fused feature vector and preset hierarchical scoring indicators. The machine learning model includes, but is not limited to, logistic regression and random forest models.

[0054] Furthermore, the machine learning model can classify the credit score according to a preset credit scoring level to obtain the corresponding credit rating. The preset credit scoring level is divided into four levels: A, B, C, and D, corresponding to quantitative scores and risk levels, as detailed below:

[0055] A credit rating of A, with a credit score ranging from 80 to 100, indicates that the company is in a low-risk credit state and signifies that the company has strong repayment ability and excellent credit status.

[0056] A credit rating of B, with a credit score ranging from 60 to 79, indicates that the company's credit is in a low to medium risk state and is used to indicate that the company has good repayment ability and no obvious abnormalities in its credit status.

[0057] A credit rating of C, with a credit score ranging from 40 to 59, indicates that the company's credit is in a medium-to-high risk state, signifying that the company has a weak repayment ability and certain credit flaws.

[0058] A credit rating of D, with a credit score ranging from 0 to 39, indicates that the company is in a high-risk state, signifying extremely poor repayment ability and serious credit problems.

[0059] In one embodiment, optionally, before step 101, the method further includes:

[0060] The operational health score is obtained based on the target robot's operational status data and a preset operational health assessment model.

[0061] In this embodiment of the invention, the operating status data is input into a preset business health assessment model. The business health assessment model outputs the business health score based on the operating status data and preset business health assessment indicators, thereby dynamically displaying the business activity status of the enterprise leasing the target robot.

[0062] Furthermore, the business health assessment model can classify the business health score according to a preset business health rating level to obtain the corresponding business health level. The preset business health rating level is divided into four levels: A, B, C, and D, corresponding to quantitative scores and risk levels, as detailed below:

[0063] The business health level is A, and the business health score ranges from 80 to 100. At this time, the enterprise is in a high-quality business state, which is used to indicate that the target robot is operating normally and the enterprise's production and operation activity is high.

[0064] The business health level is B, and the business health score ranges from 60 to 79 points. At this time, the enterprise is in a good business state, which is used to indicate that the target robot is operating normally and there is no significant decline in business activity.

[0065] The business health level is C, and the business health score ranges from 40 to 59. At this time, the enterprise is in a normal operating state, which is used to indicate that the target robot has a slight abnormality in operation and the business activity has decreased.

[0066] The business health level is D, and the business health score ranges from 0 to 39 points. At this time, the enterprise is in a poor business state, which is used to indicate that the target robot is operating abnormally and the enterprise's production and operation scale has shrunk significantly.

[0067] In one embodiment, optionally, before step 101, the method further includes:

[0068] The appraised value of the equipment is obtained based on at least one of the following from the macro market data on robots: initial value of the equipment, basic annual depreciation rate, and market conditions; the uptime and failure frequency from the operating status data of the target robot; and a preset equipment value assessment model.

[0069] In this embodiment of the invention, the equipment value assessment model is shown in the following formula (1):

[0070] V_current = V_initial × (1- _base×t)×K_working×K_market(1);

[0071] Where V_current represents the assessed value of the equipment; V_initial represents the initial value of the equipment; _base represents the base annual depreciation rate; t represents the usage time, which can be determined based on the uptime in the target robot's operating status data; K_working represents the working condition correction coefficient, which can be calculated based on the uptime and failure rate in the operating status data. Generally, the value of this working condition correction coefficient ranges from 0.3 to 1; K_market represents the market price correction coefficient, which can be calculated based on dynamic market price data. Generally, the value of this market price correction coefficient ranges from 0.4 to 1.2.

[0072] It should be noted that the assessed value of the equipment can be used as the residual value when the target robot is returned or repurchased.

[0073] In this embodiment of the invention, the equipment valuation model can be used to calculate the equipment valuation value in real time. The equipment valuation value can be referred to as the current fair value and the future predicted residual value (i.e., the residual value when the target robot is returned or repurchased).

[0074] Specifically, the equipment valuation model can obtain an equipment value score based on the ratio of the appraised value of the equipment to the outstanding lease balance of the target robot, i.e., the asset coverage ratio, including:

[0075] The asset coverage ratio is greater than or equal to 150%, the equipment value score is 90 to 100, the asset collateral is sufficient, and the value is stable;

[0076] The asset coverage ratio is less than 150% but greater than or equal to 120%, the equipment value score is 70 to 89, the asset collateral is basically sufficient, and the value fluctuates slightly.

[0077] The asset coverage ratio is less than 120%, the equipment value score is less than 70, the asset collateral is insufficient, and the value has been greatly reduced.

[0078] In one embodiment, optionally, prior to step 101, the method further includes at least two of the following:

[0079] Based on at least one of the following in the target robot's operational status data: equipment utilization rate, capacity utilization rate, failure frequency, maintenance frequency, operating load, and uptime, a risk warning result for the equipment operation dimension is obtained.

[0080] Based on the enterprise's operational data, risk warning results for the enterprise's operational dimensions are obtained;

[0081] A credit rating is obtained based on the credit score, and a risk warning result for the enterprise's credit dimension is obtained based on the credit score and / or credit rating.

[0082] Based on the assessed value of the equipment, risk warning results are obtained for the equipment value dimension.

[0083] In this embodiment of the invention, a risk warning model can be provided. This risk warning model is used to indicate: for multi-dimensional risk warning, multi-dimensional risk warning rules (or risk triggering rules, risk triggering conditions) and multi-dimensional risk warning thresholds are set, and a rule engine and anomaly detection algorithm are established to obtain multi-dimensional risk warning results, such as continuous decline in equipment operating rate, new judicial risks, deterioration of equipment health status, decline in credit score, etc., to achieve risk identification.

[0084] It should be noted that the analysis layer in the financial risk control system of the robot leasing business can provide the financial risk control intelligent agent with enterprise credit scoring models, operational health assessment models, equipment value assessment models, and risk warning models.

[0085] Furthermore, the risk warning results from at least two of the above dimensions are fused to obtain multi-dimensional risk warning results, and the multi-dimensional risk warning results are then classified into risk warning levels, specifically including:

[0086] The initial risk warning level is triggered by a single non-core indicator in at least one dimension of the multi-dimensional risk warning results (such as the number of equipment failures in a single month exceeding 1.5 times the historical level), and there is no substantial operational or credit risk. At this time, only attention is required.

[0087] Intermediate risk warning level: At least one of the risk warning results in the multi-dimensional risk warning results triggers a single core indicator threshold (such as a decrease in the operating rate of more than 30% for 7 consecutive days or a decrease in tax payment of more than 40%), and there is potential risk, so intervention is required.

[0088] Advanced risk warning level: Multiple core indicators in at least one dimension of the multi-dimensional risk warning results trigger thresholds or show significant risk signals (such as new large-scale judicial litigation, asset coverage ratio <120%, credit score down to C or below), and the risk has already manifested, requiring emergency handling.

[0089] Specifically, regarding the device operation dimension, based on the target robot's operational status data and preset risk warning rules and thresholds, the system monitors whether the target robot's operational status is abnormal. Specifically, the risk warning rules and thresholds for the device operation dimension include:

[0090] (1) Equipment operating rate: A risk warning will be triggered if any of the following conditions are met:

[0091] Equipment operating rate has decreased by more than 30% for seven consecutive days;

[0092] The daily equipment utilization rate was 50% lower than the industry average.

[0093] For 15 consecutive days, the equipment operating rate was lower than the industry benchmark of 60%.

[0094] (2) Capacity utilization rate: If any of the following conditions are met, a risk warning will be triggered:

[0095] Capacity utilization has declined by more than 25% for five consecutive days;

[0096] Capacity utilization has consistently been below 40% of the equipment's designed capacity.

[0097] (3) Fault frequency or maintenance frequency, if any of the following conditions are met, a risk warning will be triggered:

[0098] The number of outages in a single month exceeded twice the historical average.

[0099] A major malfunction (such as requiring downtime for repairs exceeding 48 hours) was not reported in a timely manner;

[0100] After three consecutive repairs, the system still could not be restored to normal operating load.

[0101] (4) The running load meets one of the following conditions, triggering a risk warning:

[0102] The average operating load of the equipment is less than 50% of the design load for 10 consecutive days;

[0103] The load fluctuation exceeds 80%.

[0104] (5) The power-on duration must meet one of the following conditions to trigger a risk warning:

[0105] For seven consecutive days, the average daily operating time of the equipment was 60% lower than the industry average.

[0106] Suddenly, the equipment had zero startup time for three consecutive days without any reasonable shutdown reporting.

[0107] Regarding the aforementioned business operations dimension, based on the company's business operation data and preset risk warning rules and thresholds, the system monitors whether risks exist in the company's fundamental business operations. Specifically, the risk warning rules and thresholds for the aforementioned business operations dimension include:

[0108] (1) Judicial litigation, meeting one of the following conditions, triggers a risk warning:

[0109] New pending judicial litigation records have been added, especially economic litigation such as loan contract disputes and sales contract disputes;

[0110] The amount involved in the lawsuit exceeds 10% of the company's registered capital;

[0111] Being listed as a person subject to enforcement and having restrictions on high-level consumption are examples of dishonest records.

[0112] (2) Business and tax authorities: If any of the following conditions are met, a risk warning will be triggered:

[0113] The company's business registration information has undergone abnormal changes (such as significant changes in legal representative, registered capital, or shareholders without explanation).

[0114] Failed to file tax returns properly for two consecutive tax periods;

[0115] The amount of tax paid decreased by more than 40% compared to the previous period without a reasonable explanation;

[0116] (3) Operating costs: A risk warning will be triggered if any of the following conditions are met:

[0117] The fact that enterprises' water and electricity consumption and the number of employees paying social security contributions have decreased by more than 30% for two consecutive months indirectly reflects the contraction of their production and operation scale.

[0118] Social security contributions have been in arrears for more than 3 months;

[0119] The non-performing asset ratio of the enterprise exceeds 5%;

[0120] The company's operating profit decreased by more than 50% compared to the previous period.

[0121] (4) Enterprise qualifications: Meeting any of the following conditions will trigger a risk warning:

[0122] The company's core business qualifications (such as industry licenses and production licenses) have expired and have not been renewed;

[0123] The administrative agency issued an administrative penalty decision (such as a business-related penalty).

[0124] For the aforementioned corporate credit dimension, based on the credit score and / or credit rating, and preset risk warning rules and risk warning thresholds, the system monitors whether the corporate credit status is declining. Specifically, the risk warning rules and risk warning thresholds for the corporate credit dimension include:

[0125] (1) Rating level: A risk warning will be triggered if any of the following conditions are met:

[0126] The credit score will drop by one level within one assessment period (30 days) (e.g., from A to B, from B to C).

[0127] (2) A risk warning will be triggered if the score meets one of the following criteria:

[0128] The credit score dropped by more than 20 points in a single instance;

[0129] The credit score has decreased by more than 30 points cumulatively over two consecutive assessment periods;

[0130] The credit score dropped below 60.

[0131] Regarding the equipment value dimension, based on the equipment assessment value and preset risk warning rules and thresholds, the system monitors whether the value of the leased asset has significantly decreased. The risk warning rules and thresholds for the equipment value dimension include:

[0132] (1) Equipment value score. A risk warning will be triggered if any of the following conditions are met:

[0133] The equipment value score dropped by more than 15 points in a single instance.

[0134] The equipment value score has decreased by more than 25 points cumulatively over 30 consecutive days;

[0135] The equipment value score dropped below 70 points;

[0136] (2) The risk warning is triggered if either of the following conditions is met: the appraised value of the equipment and the outstanding lease loan balance:

[0137] The appraised value of the equipment is less than 120% of the outstanding lease loan balance, which is considered as insufficient asset collateral coverage.

[0138] (3) The equipment's assessed value, if meeting one of the following criteria, will trigger a risk warning:

[0139] The estimated value of the target robot over the next year is expected to decrease by more than 20% compared to the previous forecast.

[0140] The error in the assessed value of the equipment exceeds the model benchmark value of ±8%, which is determined to be an abnormal fluctuation in asset value.

[0141] In one implementation, optionally, the scenario includes at least one of the following:

[0142] Pre-loan approval scenarios; daily monitoring scenarios during the loan process; risk warning triggering scenarios during the loan process.

[0143] The financial risk control decision results include at least one of the following:

[0144] Loan approval recommendations; margin ratio recommendations; lease interest rate recommendations; equipment buyback contingency plan.

[0145] Next, for different scenarios, the specific financial risk control decision results include:

[0146] Scenario 1: Pre-loan approval

[0147] Applicable stage: When a company applies for robot leasing financial services but has not yet signed a leasing agreement;

[0148] Judgment Logic: Credit score, equipment valuation, business health score (historical + current data), and risk warning results (pre-loan risk screening, no warning trigger).

[0149] The results of the financial risk control decisions are shown in Table 1 below:

[0150] Table 1

[0151]

[0152] Scenario 2: Daily monitoring during loan processing:

[0153] Applicable stage: The enterprise has signed a lease agreement, is fulfilling its repayment obligations normally, and no risk warning signals have been triggered;

[0154] Judgment Logic: Regularly (e.g., monthly) assess credit score, asset value, and business health score. All three indicators remain stable or fluctuate slightly, with no warning signals.

[0155] The results of the financial risk control decisions are shown in Table 2 below:

[0156] Table 2

[0157]

[0158] Scenario 3: Loan risk warning triggered:

[0159] Applicable stage: During the enterprise lease period, when basic / intermediate / advanced risk warning signals are triggered;

[0160] Judgment Logic: Based on the risk warning level, combined with the real-time status of credit score, equipment valuation, and business health score, decision-making rules are formulated in a tiered manner to achieve gradual risk management;

[0161] The financial risk control decision-making results include:

[0162] (1) Sub-scene, triggering a primary warning:

[0163] Judgment criteria: A single non-core indicator triggers the threshold, and the credit score, asset value, and business health score are all maintained at a normal level (credit ≥ B, asset coverage ratio ≥ 120%, business health score ≥ 60).

[0164] Loan approval recommendation: Suspend applications for additional loans; loan renewals will be reviewed normally.

[0165] Margin adjustment: unchanged;

[0166] Interest rate fluctuation: Remain unchanged;

[0167] Equipment recovery: None. The system automatically reminds management personnel to pay attention and updates indicator data daily.

[0168] (2) Sub-scene, triggering intermediate warning:

[0169] The results of the financial risk control decisions are shown in Tables 3.1 and 3.2 below:

[0170] Table 3.1

[0171]

[0172] Table 3.2

[0173]

[0174] (3) Sub-scene, trigger advanced alert:

[0175] The results of the financial risk control decisions are shown in Table 4 below:

[0176] Table 4

[0177]

[0178] Furthermore, after step 101, the embodiment of the present invention further includes:

[0179] The multi-dimensional risk warning results are processed in a tiered manner through a tiered early warning system to achieve progressive risk warning, including:

[0180] For the initial risk warning level, the tiered warning processing includes message notifications;

[0181] For the intermediate risk warning level, the tiered warning processing includes pushing the multi-dimensional risk warning results to the management personnel terminal;

[0182] For high-risk warning levels, the tiered warning process includes audible and visual alarms and special reports.

[0183] Furthermore, after step 101, this embodiment of the invention further includes:

[0184] The automated strategy execution unit can trigger corresponding financial risk control strategies based on the financial risk control decision results, such as sending a deposit replenishment notice to the company leasing the target robot, adjusting the leasing interest rate, and initiating the equipment recycling process.

[0185] The report generation unit regularly generates credit scoring reports, equipment valuation reports, and risk monitoring reports, providing financial risk decision-making references for funders and managers.

[0186] It should be noted that, prior to step 101, the embodiments of the present invention further include:

[0187] Based on the raw data, model calculation data, decision result data, risk warning result data, financial risk control strategy execution data, and risk warning processing effect data of the entire process of pre-loan, mid-loan, and post-loan, a standardized risk control learning dataset is constructed.

[0188] Based on the risk control learning dataset, the parameters of the core models mentioned above, such as the enterprise credit scoring model, business health assessment model, equipment value assessment model, and risk warning model, are iterated and the structure is optimized. The model weights and feature dimensions are dynamically adjusted to improve the model prediction and recognition accuracy.

[0189] The financial risk control intelligent agent autonomously optimizes risk warning thresholds, financial risk control decision-making results, and warning classification rules based on historical risk warning processing effect data and robot market environment change data, making the rule system more adaptable to actual risk control scenarios.

[0190] The financial risk control intelligence agent puts the optimized model and rules into the test set for verification. If the verification effect reaches the preset standard (such as improved risk identification accuracy and reduced false judgment rate), the optimized model and rules are autonomously put into actual use. If the standard is not met, the learning and optimization are carried out again.

[0191] In one implementation, optionally, obtaining the financial risk control decision results for different scenarios output by the financial risk control intelligent agent based on the input state in step 101 above includes:

[0192] Obtain the financial risk control decision results for different scenarios output by the financial risk control intelligent agent based on the manual intervention interface module and the input state; wherein, the manual intervention interface module includes:

[0193] The manual review interface allows administrators to manually review and adjust the financial risk control decision results, which is suitable for special clients or special risk control scenarios.

[0194] The rule adjustment interface allows administrators to manually adjust the risk warning rules and risk warning thresholds of the financial risk control intelligence agent according to market changes or business needs. The financial risk control intelligence agent will record the adjustments and use them as a basis for self-learning.

[0195] The manual risk event handling interface allows managers to manually handle sudden risk events. The financial risk control intelligent agent will cooperate to execute financial risk control strategies and record the process and effects of manual handling.

[0196] In one embodiment, optionally, before step 101, the method further includes:

[0197] For at least two of the following dimensions—equipment operation dimension, enterprise operation dimension, enterprise credit dimension, and equipment value dimension—a sliding time window model corresponding to the risk warning conditions is configured. The sliding time window model includes at least one of the following: window length, sliding step size, and data sampling frequency.

[0198] Based on the sliding time window model, acquire the enterprise's business data, the target robot's operating status data, and the robot's macro market data.

[0199] In this embodiment, a risk warning based on a sliding time window can be implemented. A configurable sliding time window model is designed for the risk indicator characteristics of four dimensions: equipment operation, enterprise management, enterprise credit, and equipment value. It supports personalized settings for window length, sliding step size, and data sampling frequency, ensuring compatibility with the monitoring cycle of each dimension's risk warning conditions, as detailed below:

[0200] (1) Core configuration parameters:

[0201] Window length (T): Values ​​range from 7 to 30 days, with default values ​​preset according to risk indicator type, and manual customization is supported;

[0202] Sliding step size (S): This is the time interval between each time the window slides forward. It is divided into second / minute level (equipment operation indicators) and hour / day level (operation / asset / credit indicators) according to the data collection frequency.

[0203] Data sampling frequency (F): Synchronized with the upstream data acquisition unit, equipment operation indicators are collected at the second level, and enterprise operation / asset / credit indicators are collected at the daily / weekly / monthly level;

[0204] Window data storage strategy: Use scrolling window storage, retain only the data that overlaps between the current window and the previous window, remove expired data, and reduce storage costs.

[0205] (2) Default values ​​for multi-dimensional window configuration:

[0206] Based on the risk warning conditions set by the financial risk control intelligent agent, differentiated sliding time windows are configured for the four dimensions of indicators, matching the time judgment requirements of the thresholds, as shown in Table 5 below:

[0207] Table 5

[0208]

[0209] (3) Calculation and implementation of the sliding time window:

[0210] The core logic for implementing a scrolling sliding window based on a real-time computing engine (such as Flink) is as follows:

[0211] The multi-source data preprocessed upstream is fed into the real-time computing engine in the form of a data stream, and each data point is timestamped and labeled with indicator dimensions.

[0212] By matching preset sliding time window configuration parameters according to indicator dimensions, a dedicated window calculation task is constructed.

[0213] The window scrolls forward in a set step size. Each time it scrolls, it automatically captures all valid data in the window and removes expired data whose timestamps exceed the window length.

[0214] Aggregate and calculate the data captured within the window to obtain statistical values ​​of the indicators (such as mean, cumulative value, rate of change, month-on-month / year-on-year decrease, etc.), providing a data foundation for subsequent threshold matching;

[0215] The window calculation results are pushed to the indicator database and rule engine in real time, while the original data is retained to support risk traceability.

[0216] For example, in equipment utilization rate monitoring, the risk warning condition is a decrease in utilization rate of more than 30% for 7 consecutive days;

[0217] Window configuration: 7-day length, 1-hour sliding step, and second-level sampling frequency;

[0218] Calculation logic: The window is scrolled once every hour to capture the second-level data of the operating rate for the past 7 days. The average daily equipment operating rate and the cumulative change rate of the equipment operating rate over 7 days are calculated. If the cumulative change rate is ≤-30%, the threshold matching is triggered.

[0219] (4) Early warning system combining multi-dimensional and sliding time window:

[0220] This approach deeply integrates dynamic data calculations from sliding time windows with risk warning threshold matching based on multi-dimensional risk warning conditions. Combined with the isolated forest anomaly detection algorithm, it achieves dual warnings: direct triggering of explicit risks and early identification of implicit risks. The entire process consists of 7 core steps, executed fully automatically with a total response time of less than 5 minutes, as detailed below:

[0221] Step 1: Real-time access and preprocessing of multi-source data:

[0222] The upstream multi-source data acquisition module cleans, deduplicates, and fills in missing values ​​for enterprise operation data, operational status data, robot macro market data, credit scoring data, and equipment value assessment data.

[0223] Step 2, Dynamic data slicing and aggregation calculation based on sliding time windows:

[0224] The real-time computing engine automatically matches preset sliding time window configuration parameters (such as window length, sliding step size, and data sampling frequency) based on the data's indicator dimension labels.

[0225] The standardized data stream is dynamically sliced ​​according to the sliding time window rule, and all valid data in the current window is extracted.

[0226] The sliced ​​data is aggregated and calculated to obtain the statistical values ​​of the indicators required for risk monitoring. The core calculation types include:

[0227] Change rate: such as the cumulative change rate of the 7-day operating rate, or the change rate of the 30-day credit score.

[0228] Cumulative / Average: such as the number of equipment failures in a single month, or the average daily uptime;

[0229] Ratio / Percentage: such as asset coverage ratio (fair value of equipment / outstanding loan balance), and the percentage decrease in tax payment compared to the previous period;

[0230] Threshold comparison values: such as the difference between the operating rate and the industry average, or the ratio of capacity utilization rate to the designed capacity of equipment.

[0231] The calculation results are pushed to the rule engine and indicator database in real time, while the original data and calculation process are retained.

[0232] Step 3: Multi-dimensional risk warning threshold matching by the rule engine:

[0233] The rule engine retrieves the risk trigger rules for the corresponding dimension from the rule base and accurately matches the indicator calculation results of the sliding time window with the risk warning threshold one by one;

[0234] For single indicator rules, directly determine whether the trigger logic is met. For example, if the capacity utilization rate decreases by more than 25% for 5 consecutive days, and the cumulative change rate of the capacity utilization rate calculated by the sliding time window is ≤-25%, then the risk warning rule is triggered.

[0235] For combined indicator rules, such as advanced alerts involving two or more core indicator dimensions, the rule matching results of multiple dimensions are checked, and the combined rule is triggered if the "AND / OR" logic is satisfied.

[0236] If a match is successful, an explicit risk trigger signal is generated, marking the trigger level (basic / intermediate / advanced), trigger dimension, and association rule; if a match fails, proceed to step 4 for implicit risk anomaly detection.

[0237] Step 4, Latent risk anomaly detection in the Isolation Forest algorithm:

[0238] For cases where the explicit threshold is not triggered but the indicator trend is abnormal, the Isolation Forest anomaly detection algorithm is used to analyze the indicator data trend within the sliding time window to identify potential hidden risks, including the following steps:

[0239] Using time series data of indicators within a sliding time window as training samples (such as asset value score time series data of the past 30 days), an isolated forest model is constructed, and an abnormal score threshold is set (by default, an abnormal score ≥0.7 is considered abnormal).

[0240] The model calculates the outlier of the time series data of the indicators in the current window and identifies outliers and abnormal trends in the data (such as a small but continuous decline in equipment utilization rate without reasonable cause, and non-periodic fluctuations in asset value scores).

[0241] If the abnormal score is greater than or equal to the preset threshold, a hidden risk warning signal will be generated, and it will be marked as a primary warning, indicating that there is a potential risk, and pushed to the management personnel for manual verification.

[0242] If the abnormal score is less than the preset threshold, the indicator is considered normal, the current monitoring process ends, and the cycle of the next sliding window begins.

[0243] Step 5, Risk warning result classification determination:

[0244] By combining explicit risk warning results with implicit risk warning results, and based on the warning classification rules (basic / intermediate / advanced) set by the financial risk control intelligent agent, the risk warning results are automatically classified. The core judgment logic is as follows:

[0245] Early warning: A single non-core indicator triggers explicit rules or the isolated forest algorithm detects hidden risks (such as no substantial operational / credit risks).

[0246] Intermediate warning: A single core indicator triggers an explicit rule (such as a decrease in the operating rate of more than 30% for 7 consecutive days or a decrease in tax payment of more than 40%), while no other dimensional indicators trigger it;

[0247] Advanced warning: Multiple core indicators trigger explicit rules or show significant risk signals (such as new large-scale litigation, asset coverage ratio <120%, credit score down to C or below).

[0248] The risk warning results after classification will be accompanied by complete judgment criteria, including indicator calculation values, threshold ranges, trigger rules, and anomaly scores (if any), to ensure that decisions are traceable.

[0249] Step 6: Multi-channel output and recording of risk warning results:

[0250] The risk warning results after classification are output to the classified warning system of the intelligent decision-making and execution module in real time, and all information is recorded at the same time, as follows:

[0251] Signal output: The corresponding output channels are matched according to the warning level. The primary warning is a system message prompt, the intermediate warning is a push notification to the management personnel terminal, and the advanced warning is an audible and visual alarm and automatic generation of special reports.

[0252] Information Recording: All core information of the risk warning results (such as classification results, triggering dimensions, indicator data, judgment basis, and output time) are recorded into the risk warning log library and permanently stored to support subsequent risk tracing and model optimization;

[0253] Data synchronization: The risk warning results are synchronized to the risk control decision engine to provide a basis for the execution of downstream automated financial risk control strategies.

[0254] Step 7, Tracking and Closed-Loop Processing of Risk Warning Results:

[0255] For the risk warning results that have been triggered, the system will start continuous tracking and monitoring, continue to collect indicator data according to the original sliding time window configuration, and update the indicator change trend in real time.

[0256] If the indicators return to normal (such as the equipment operating rate rising above the threshold or the enterprise submitting supplementary tax returns), the warning will be automatically lifted, and the lifting time and the reason for the recovery will be recorded.

[0257] If the indicators continue to deteriorate (such as the initial warning being upgraded to an intermediate warning, or the intermediate warning triggering multiple core rules), the warning level will be automatically upgraded, and stricter financial risk control strategies will be triggered downstream.

[0258] It supports manual intervention, allowing managers to manually verify, mark, cancel, or adjust risk warning rules and sliding time window configuration parameters through the manual intervention interface, ensuring the flexibility of financial risk control.

[0259] Figure 2 This is a flowchart illustrating one embodiment of the financial risk control method for robot leasing business described in this invention. Figure 2 As shown, the method includes:

[0260] Phase 1, Lessee Company:

[0261] To apply for robot rental business, submit your company information;

[0262] After signing the lease agreement, the robot was received and put into use;

[0263] Make timely repayments during the robot rental period.

[0264] Phase Two, Financial Risk Control Intelligent Agent:

[0265] Collect data from multiple sources;

[0266] Based on multi-source data, credit scores, equipment assessment values, and business health scores are obtained.

[0267] Risk control decision results are obtained through the distributed control decision engine;

[0268] Check the results of risk control decisions, initiate real-time post-loan monitoring, and collect operational status data in real time.

[0269] Phase Three, Risk Warning:

[0270] Calculate the sliding time window and match the risk warning rules with the risk warning threshold. Figure 2 (abbreviated as matching rule threshold) is used in conjunction with isolated forest anomaly detection to determine the risk warning level;

[0271] Determine whether a risk warning has been triggered based on the risk warning level;

[0272] If no risk warning is triggered, monitoring will continue; if a risk warning is triggered, a tiered warning will be issued.

[0273] Phase Four, Strategy Execution:

[0274] Implement financial risk control strategies and generate financial risk reports;

[0275] Handle according to different levels of early warning;

[0276] For initial or intermediate warning responses, determine whether the indicators have recovered;

[0277] If the indicators recover, the warning will be lifted; if the indicators do not recover, the warning level will be upgraded.

[0278] Phase 5, Agent Learning:

[0279] Data accumulation, model and rule optimization;

[0280] Verify the learning outcomes and determine whether the verification is successful.

[0281] If the verification passes, the financial risk control intelligent agent can be put into production; if the verification fails, the financial risk control intelligent agent will be re-optimized.

[0282] Figure 3 This is a schematic diagram of the architecture of the application system for the financial risk control method in robot leasing business according to an embodiment of the present invention. Here, the application system for the financial risk control method in robot leasing business can be referred to as the financial risk control system for robot leasing business, hereinafter simply referred to as the system. Figure 3 As shown, the system includes:

[0283] The perception layer provides multi-dimensional data input to the financial risk control intelligent agent, enabling the financial risk control intelligent agent to have a comprehensive data perception of enterprise operations, equipment operation, and macro market. All collected and pre-processed data are transmitted to the core data platform of the financial risk control intelligent agent for unified management and scheduling.

[0284] The analysis layer is the core algorithm support of the financial risk control intelligent agent. All models are uniformly scheduled and run by the financial risk control intelligent agent, and the model calculation results are fed back to the decision center of the financial risk control intelligent agent in real time, providing algorithmic basis for the decision-making of the financial risk control intelligent agent.

[0285] The decision-making and execution layer is the core decision-making and action output component of the financial risk control intelligent agent. Based on the model results of the analysis layer, the financial risk control intelligent agent autonomously schedules and completes risk control decisions, early warnings, strategy execution, and report generation. All actions are autonomously triggered and completed by the financial risk control intelligent agent.

[0286] The learning layer is the core component for the financial risk control intelligence to achieve autonomous iterative upgrades. It is autonomously scheduled and operated by the financial risk control intelligence, and completes self-learning and model optimization based on data from the entire risk control process to achieve continuous improvement in risk control capabilities.

[0287] The human-machine collaboration layer includes the aforementioned manual review interface, rule adjustment interface, and manual handling interface for risk events.

[0288] In summary, the financial risk control method for robot leasing business described in this invention achieves intelligentization. With a financial risk control intelligent agent at its core, it constructs a six-layer integrated architecture encompassing perception, analysis, decision-making, execution, learning, and human-machine collaboration. This overcomes the shortcomings of existing financial risk control systems where each module operates independently, enabling autonomous perception, analysis, decision-making, execution, and learning throughout the entire financial risk control process. It solves the problems of insufficient comprehensiveness and accuracy in financial risk control, while also considering flexible human-machine collaboration capabilities. This upgrades the financial risk control system from a modular algorithm combination to a financial risk control intelligent agent with autonomous intelligence, significantly improving the efficiency and response speed of the financial risk control process and achieving an intelligent closed loop for the entire financial risk control process.

[0289] Moreover, credit scoring is more accurate and comprehensive, combining comprehensiveness and accuracy in risk control decisions. Specifically, automated strategies cover the entire process of pre-loan, during-loan, and post-loan; flexible adjustment is supported through human-machine collaboration. While achieving autonomous operation throughout the entire process, the financial risk control intelligence provides an interface for human intervention. Humans can adjust the decisions and rules of the financial risk control intelligence in special scenarios. The financial risk control intelligence will record all human operations and incorporate them into its self-learning, achieving the optimal financial risk control model with intelligent autonomy as the main approach and human assistance as a supplement.

[0290] The financial risk control intelligence agent has the ability to learn and iterate autonomously, accumulate risk control data throughout the entire process, and learn from historical data, decision results, and risk handling effects. It can autonomously optimize model parameters, risk warning thresholds, and risk warning rules without the need for manual model iteration and rule adjustment. This allows the risk identification accuracy and decision-making precision of the financial risk control intelligence agent to continuously improve financial risk control capabilities as business operations progress.

[0291] And, it has the following significant application effects: (1) the accuracy of risk control is significantly improved and the misjudgment rate is reduced by 35%; (2) the real-time response capability is strong, the data update delay and risk warning response are significantly reduced, and potential risks can be dealt with quickly; (3) the degree of automation of the whole process is high, reducing the workload of manual review by more than 80%, improving risk control efficiency and reducing operating costs; (4) the degree of intelligence is high, using the large model development capability, all data are used as prompt words, and the financial risk control intelligent agent makes autonomous decisions, and then the financial risk control decision results are intelligently sent in different forms such as text, sound, and message body to realize the intelligent decision-making and notification; financial risk control intelligence The intelligent agent is designed for robot leasing financial scenarios and can flexibly adapt to the leasing risk control needs of different types of robots (such as industrial robots, service robots, etc.). It can also adapt to the risk control characteristics of different regions and industries through self-learning; (6) Data compliance and security: All data collection, transmission and storage of the financial risk control intelligent agent adopt encryption technology, strictly follow data security regulations, protect sensitive information of enterprises and funders, and the entire data operation is uniformly controlled by the intelligent agent to reduce the risk of data leakage; (7) Transparent and traceable decision basis: The financial risk control intelligent agent can permanently retain the data of the entire process, support traceability query, and can generate traceability reports independently.

[0292] Reference Figure 4 This invention also provides a financial risk control device for robot leasing business, comprising:

[0293] The control module 401 is used to determine the operational health score of the target robot, the equipment evaluation value, and the multi-dimensional risk warning results, as well as the credit score of the enterprise leasing the target robot, as the input state of the financial risk control intelligent agent, and to obtain the financial risk control decision results for different scenarios output by the financial risk control intelligent agent based on the input state.

[0294] The multi-dimensional risk warning results include risk warning results in at least two of the following dimensions: equipment operation dimension, enterprise operation dimension, enterprise credit dimension, and equipment value dimension.

[0295] Optionally, the financial risk control device for the robot leasing business further includes at least two of the following:

[0296] The first acquisition module is used to obtain risk warning results for the equipment operation dimension based on at least one of the following in the target robot's operation status data: equipment utilization rate, capacity utilization rate, failure frequency, maintenance frequency, operating load, and startup duration.

[0297] The second acquisition module is used to obtain the risk warning result of the enterprise's business dimension based on the enterprise's business operation data;

[0298] The third acquisition module is used to obtain a credit rating based on the credit score, and to obtain a risk warning result for the enterprise credit dimension based on the credit score and / or credit rating.

[0299] The fourth acquisition module is used to obtain risk warning results for the equipment value dimension based on the equipment assessment value.

[0300] Optionally, the financial risk control device for the robot leasing business further includes:

[0301] The configuration module is used to configure a sliding time window model corresponding to the risk warning conditions for at least two of the following dimensions: the equipment operation dimension, the enterprise operation dimension, the enterprise credit dimension, and the equipment value dimension. The sliding time window model includes at least one of the following: window length, sliding step size, and data sampling frequency.

[0302] The acquisition module is used to acquire the enterprise's business data, the target robot's operating status data, and the robot's macro market data based on the sliding time window model.

[0303] Optionally, the financial risk control device for the robot leasing business further includes:

[0304] The fifth acquisition module is used to obtain the operational health score based on the target robot's operational status data and a preset operational health assessment model.

[0305] Optionally, the financial risk control device for the robot leasing business further includes:

[0306] The sixth acquisition module is used to obtain the appraised value of the equipment based on at least one of the initial value of the equipment, the basic annual depreciation rate and market conditions data in the macro market data of the robot, the uptime and failure frequency in the operating status data of the target robot, and a preset equipment value assessment model.

[0307] Optionally, the financial risk control device for the robot leasing business further includes:

[0308] The seventh acquisition module is used to convert the enterprise's business data and the target robot's operating status data into feature vectors respectively, and to perform a weighted summation of the converted feature vectors to obtain a fused feature vector;

[0309] The eighth module is used to obtain the credit score based on the fused feature vector and a preset machine learning model.

[0310] It should be noted that the financial risk control device for robot leasing business provided in this embodiment of the invention can execute the above-mentioned financial risk control method for robot leasing business. Therefore, all embodiments of the above-mentioned cost analysis method are applicable to the financial risk control device for robot leasing business and can achieve the same or similar technical effects.

[0311] This invention also provides a financial risk control device for robot leasing businesses, such as... Figure 5 As shown, it includes:

[0312] The processor 501, memory 502, transceiver 503, and programs or instructions stored in the memory 502 and executable on the processor 501; when the processor 501 executes the programs or instructions, it implements the various processes of the above-described embodiment of the financial risk control method for robot leasing business and achieves the same technical effect. To avoid repetition, these will not be described again here.

[0313] The transceiver 503 is used to receive and send data under the control of the processor 501.

[0314] Among them, Figure 5 In this context, the bus architecture can include any number of interconnected buses and bridges, specifically connecting various circuits of one or more processors represented by processor 501 and memory represented by memory 502. The bus architecture can also connect various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. The bus interface provides an interface. Transceiver 503 can be multiple elements, including transmitters and receivers, providing a unit for communicating with various other devices over a transmission medium. For different user equipment, the user interface 504 can also be an interface capable of connecting external or internal devices, including but not limited to keypads, displays, speakers, microphones, joysticks, etc.

[0315] The processor 501 is responsible for managing the bus architecture and general processing, while the memory 502 can store the data used by the processor 501 when performing operations.

[0316] This invention also provides a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements the various processes of the above-described embodiments of the financial risk control method for robot leasing business, achieving the same technical effects. To avoid repetition, these details are not repeated here. The computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, etc.

[0317] This invention also provides a computer program product, including computer instructions. When the computer instructions are executed by a processor, they implement the various processes of the above-described embodiment of the financial risk control method for robot leasing business and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0318] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0319] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0320] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of the present invention.

Claims

1. A method for controlling financial risks in robot leasing business, characterized in that, include: The operational health score of the target robot, the equipment assessment value, and the multi-dimensional risk warning results, as well as the credit score of the enterprise leasing the target robot, are determined as the input state of the financial risk control intelligent agent, and the financial risk control decision results for different scenarios are obtained by the financial risk control intelligent agent based on the input state. The multi-dimensional risk warning results include risk warning results in at least two of the following dimensions: equipment operation dimension, enterprise operation dimension, enterprise credit dimension, and equipment value dimension.

2. The method according to claim 1, characterized in that, The method further includes at least two of the following: Based on at least one of the following in the target robot's operational status data: equipment utilization rate, capacity utilization rate, failure frequency, maintenance frequency, operating load, and uptime, a risk warning result for the equipment operation dimension is obtained. Based on the enterprise's operational data, risk warning results for the enterprise's operational dimensions are obtained; A credit rating is obtained based on the credit score, and a risk warning result for the enterprise's credit dimension is obtained based on the credit score and / or credit rating. Based on the assessed value of the equipment, risk warning results are obtained for the equipment value dimension.

3. The method according to claim 1, characterized in that, The method further includes: For at least two of the following dimensions—equipment operation dimension, enterprise operation dimension, enterprise credit dimension, and equipment value dimension—a sliding time window model corresponding to the risk warning conditions is configured. The sliding time window model includes at least one of the following: window length, sliding step size, and data sampling frequency. Based on the sliding time window model, acquire the enterprise's business data, the target robot's operating status data, and the robot's macro market data.

4. The method according to claim 1, characterized in that, The method further includes: The operational health score is obtained based on the target robot's operational status data and a preset operational health assessment model.

5. The method according to claim 1, characterized in that, The method further includes: The appraised value of the equipment is obtained based on at least one of the following from the macro market data on robots: initial value of the equipment, basic annual depreciation rate, and market conditions; the uptime and failure frequency from the operating status data of the target robot; and a preset equipment value assessment model.

6. The method according to claim 1, characterized in that, The method further includes: The enterprise's business data and the target robot's operating status data are respectively converted into feature vectors, and the converted feature vectors are weighted and summed to obtain a fused feature vector. The credit score is obtained based on the fused feature vector and the preset machine learning model.

7. A financial risk control device for robot leasing business, characterized in that, include: The control module is used to determine the operational health score of the target robot, the equipment assessment value, and the multi-dimensional risk warning results, as well as the credit score of the enterprise leasing the target robot, as the input state of the financial risk control intelligent agent, and to obtain the financial risk control decision results for different scenarios output by the financial risk control intelligent agent based on the input state. The multi-dimensional risk warning results include risk warning results in at least two of the following dimensions: equipment operation dimension, enterprise operation dimension, enterprise credit dimension, and equipment value dimension.

8. A financial risk control device for robot leasing business, characterized in that, include: A processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the processor, when executing the program or instructions, implements the financial risk control method for robot leasing business as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the financial risk control method for robot leasing business as described in any one of claims 1 to 6.

10. A computer program product, characterized in that, It includes computer instructions, which, when executed by a processor, implement the financial risk control method for robot leasing business as described in any one of claims 1 to 6.

Images