Order management method and system based on ERP

Monitor the abnormal state of ERP and CRM systems through machine learning algorithms and rule methods, and combine one-way encryption verification and logistic regression model evaluation to solve the data transmission problem of ERP and CRM systems under high load conditions, ensuring data credibility and system stability avoid data loss or distortion caused by system crashes.

CN119006094AActive Publication Date: 2024-11-22JIANGSU SUZHONG PRINTING CO LTD

Patent Information

Application Number
CN202411011206.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-26
Publication Date
2024-11-22
Estimated Expiration
2044-07-26

AI Technical Summary

Technical Problem

Existing ERP and CRM systems have problems such as data consistency, process specifications, permission management and system performance during data transmission, resulting in system errors or crashes under high load conditions, resulting in the failure of ERP system order management and it is difficult to determine the failure. source.

Method used

The abnormal state of the ERP system is monitored by machine learning algorithm, combined with rules-based methods to monitor the abnormal state of the CRM system, and verifies the trustworthiness of the data through one-way encryption, uses logistic regression model to evaluate the order data transmission process, and formulates the Optimize strategies to ensure the trustworthiness of data transmission and system stability.

Benefits of technology

By comprehensively evaluating the status of ERP and CRM systems, ensuring the credibility of data transmission and avoiding data loss or distortion caused by system crashes, it provides optimization strategies in the event of system failure to ensure the security and credibility of order data.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119006094A_ABST
    Figure CN119006094A_ABST
Patent Text Reader

Abstract

The invention discloses an order management method and system based on ERP (Enterprise Resource Planning), and particularly relates to the technical field of order management, and the method specifically comprises the steps: carrying out the periodic inspection of the stability of an ERP system and a CRM (Customer Resource Management) system, carrying out the monitoring, evaluation and early warning of the abnormal states of the ERP system and the CRM system, and when the ERP system or the CRM system enters an early warning state, carrying out the early warning of the ERP system or the CRM system. The method comprises the following steps: verifying order data transmitted from a CRM system to an ERP system, checking data credibility through a one-way encryption mode, evaluating a transmission process of the order data by applying a logistic regression model and combining load states of the ERP system and the CRM system, and performing strategy optimization on order management according to an evaluation result of the logistic regression model. The transmission credibility of the original data among different systems is effectively controlled, the abnormal fault link of the original data is positioned according to the logistic regression model, and the problem of data loss or distortion caused by single-system or dual-system crash is avoided.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of order management, and more specifically, to an order management method and system based on ERP. Background Art

[0002] Existing enterprise order management is divided into two parts due to differences in functional scope, which are composed of internal ERP, i.e. enterprise resource planning, and external CRM, i.e. customer relationship management. The core of the ERP system is to integrate various business processes within the enterprise, including financial management, supply chain management, human resource management, manufacturing management, project management, etc., to improve internal operational efficiency, optimize resource allocation, and reduce costs. The CRM system focuses on managing and analyzing customer information and interactions, including sales automation, marketing automation, customer service and support, etc., to improve customer satisfaction, increase sales revenue, and optimize customer relationships. In the intersection of the two, they collaborate through shared data and directly introduce order information in the CRM system into the ERP system for processing. However, during the circulation process, information belonging to the two different systems is still restricted by data consistency, process specifications, authority management, and system performance between different systems. Once the order data based on CRM generates system errors or crashes under high load and large traffic conditions, it may cause the failure of ERP system order management, and it is difficult to determine the exact link of the root cause of the failure.

[0003] In order to solve the above defects, a technical solution is now proposed. Summary of the invention

[0004] The purpose of the present invention is to provide an order management method and system based on ERP to solve the problems in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solution: an order management method based on ERP, the specific steps of which include periodically checking the stability of the ERP system and the CRM system, monitoring the abnormal state of the ERP system by a method based on a machine learning algorithm, and monitoring the abnormal state of the CRM system by a rule-based method;

[0006] Evaluate and warn of abnormal status of ERP and CRM systems. When the ERP or CRM system enters the warning state, verify the order data transmitted from the CRM system to the ERP system and check the credibility of the data through one-way encryption.

[0007] Use logistic regression model to evaluate the order data transmission process in combination with the load status of ERP system and CRM system;

[0008] Based on the evaluation results of the logistic regression model, the order management strategy is optimized.

[0009] Preferably, the method for periodically checking the stability of the ERP system and the CRM system is:

[0010] Set up a system monitoring tool or framework for heartbeat detection, configure the monitored system components and set the frequency of sending heartbeat signals;

[0011] Write scripts to send heartbeat signals to the ERP system and CRM system regularly, receive the heartbeat signals through system monitoring tools and set thresholds;

[0012] If the heartbeat signal is not received within the set time, an early warning is triggered.

[0013] Preferably, a method based on a machine learning algorithm is used to monitor the abnormal state of the ERP system, and the specific method is:

[0014] Collect log data, order records, operation records, and performance indicators of the ERP system;

[0015] Perform data preprocessing on the collected log data, order records, operation records, and performance indicators, remove or correct missing values, noise data, and outliers, and normalize the preprocessed data;

[0016] The time series features of normalized log data, order records, operation records, and performance indicators are used as the input of the feature vector. The time series data is divided into windows of fixed length using the sliding window method. The window numbers are marked as w = {1, 2, ... k}, where k is the number of windows and k is a positive integer, and the features of each window are extracted.

[0017] The extracted feature vector is input into the long short-term memory network model for training and the model output value is recorded. The model output value is marked as L op .

[0018] Preferably, the method used for normalization is Euclidean distance normalization:

[0019] For each eigenvector x=[x1,x2,…,x n ], where n is the number of elements in the eigenvector, and n is a positive integer, there exists In the formula, v i is the normalized eigenvector, x i is the eigenvector before normalization.

[0020] Preferably, the abnormal state of the CRM system is monitored by a rule-based method, and the specific method is:

[0021] Collect CRM system indicator extraction time data through system monitoring tools, including system log retrieval time, error rate calculation time, query response test time, and perform data preprocessing;

[0022] According to the window segmentation method of the sliding window, the system log retrieval time, error rate calculation time, and query response test time are marked and numbered, and the system log retrieval time is calibrated as S w The error rate calculation takes C w , the query response test takes T w , where w is the sequence number of the sliding window;

[0023] The rule model for configuring the system monitoring tool is: Where Ad is the rule response index, α and β are ln(S w +1) and The weight coefficient of , and α and β are both positive integers.

[0024] Preferably, the logic for integrated evaluation and early warning of abnormal status of the ERP system and the CRM system is:

[0025] The classification threshold for the long short-term memory network model is preset to C t , the abnormal threshold of the rule model is preset as E t , when the calculated model output value L of the long short-term memory network model op Greater than or equal to the classification threshold C t , and the calculated rule response index Ad is greater than or equal to the abnormal threshold E t When the ERP system and the CRM system that transmits order data to the ERP system both experience abnormal failures, they are marked as level 1 abnormal warnings and send warning signals to management personnel;

[0026] When the model output value L of the long short-term memory network model is calculated op Less than the classification threshold C t , and the calculated rule response index Ad is greater than or equal to the abnormal threshold E t When the ERP system does not have any abnormal failure, but the CRM system that transmits order data to the ERP system has an abnormal failure, which is marked as a 2A level abnormal warning and sends a warning signal to the management personnel;

[0027] When the model output value L of the long short-term memory network model is calculated op Greater than or equal to the classification threshold C t , and the calculated rule response index Ad is less than the abnormal threshold E tWhen the ERP system has an abnormal failure, but the CRM system that transmits order data to the ERP system has no abnormal failure, it is marked as a 2B level abnormal warning and a warning signal is sent to the management personnel;

[0028] When the model output value L of the long short-term memory network model is calculated op Less than the classification threshold C t , and the calculated rule response index Ad is less than the abnormal threshold E t When , there is no abnormal failure in the ERP system and the CRM system that transmits order data to the ERP system, it is marked as normal status and the normal signal is recorded.

[0029] Preferably, the verification method for verifying the credibility of data by one-way encryption is:

[0030] The CRM system converts the benchmark order data into a string or binary format and sends it to the ERP system. The CRM system also uses a hash algorithm to calculate a benchmark hash value based on the string or binary format of the benchmark order data. The benchmark hash value is then saved by the ERP system.

[0031] When the ERP system receives the order data from the CRM system, it calculates the hash value of the order data according to the hash algorithm and verifies the calculated hash value with the reference hash value. When the verification is consistent, the data verification is marked as passed and the verification value C is set. v When the verification is inconsistent, the data verification is marked as failed and the verification value C is set. v is 0.

[0032] Preferably, the logic of using the logistic regression model to evaluate the transmission process of order data in combination with the load status of the ERP system and the CRM system is:

[0033] The calibrated resource load rate is L r , where r is the number of servers with different loads, and r = 1, 2, ... u}, where u is the number of load servers, and u is a positive integer, and the component load rate is T p , where p is the number of different computing power components in the same load server, and p = 1, 2, ... q}, where q is the number of computing power components in the load server, and q is a positive integer. A logistic regression model is established to calculate the transmission load index T rs , the calculation expression is

[0034] Preferably, the order management strategy is optimized according to the evaluation results of the logistic regression model;

[0035] Preset the transmission critical threshold T for the logistic regression model al, when the manager receives a Level 1 abnormality warning and the calculated transmission load index T rs Greater than the preset transmission critical threshold T al When the ERP system, CRM system and order data input are abnormal, the order data is unreliable, and the notification optimization strategy is to restart the ERP system and CRM system and manually input the order data;

[0036] When the manager receives a Level 1 abnormality warning and the calculated transmission load index T rs Less than or equal to the preset transmission critical threshold T al When the ERP system and CRM system are abnormal, the output of the order data is abnormal, and there is a risk of channel distortion of the order data. The notification optimization strategy is to restart the ERP system and CRM system and repeat the verification of the order data;

[0037] When the manager receives a Level 2A abnormal warning and the calculated transmission load index T rs Greater than the preset transmission critical threshold T al When the CRM system and order data input are abnormal, there is a risk of source distortion of order data. The notification optimization strategy is to restart the CRM system and manually check the order source information;

[0038] When the manager receives a Level 2A abnormal warning and the calculated transmission load index T rs Less than or equal to the preset transmission critical threshold T al When the output of the CRM system and order data are abnormal, there is a risk of channel distortion of the order data. The notification optimization strategy is to restart the CRM system and repeat the verification of the order data;

[0039] When the manager receives a Level 2B abnormality warning and the calculated transmission load index T rs Greater than the preset transmission critical threshold T al When the ERP system and order data input are abnormal, there is a risk of source distortion of order data. The notification optimization strategy is to restart the ERP system and manually check the order source information;

[0040] When the manager receives a Level 2B abnormality warning and the calculated transmission load index T rs Less than or equal to the preset transmission critical threshold T al When , the output of the ERP system and order data are abnormal, and there is a risk of channel distortion of the order data. The notification optimization strategy is to restart the ERP system and repeat the verification of the order data.

[0041] The ERP-based order management system includes a periodic verification module, a machine learning module, a rule module, a fusion evaluation module, a credibility verification module, a transmission load module, and a strategy optimization module;

[0042] The periodic verification module is used to periodically check the stability of the ERP system and CRM system;

[0043] The machine learning module is used to monitor the abnormal status of the ERP system according to a method based on a machine learning algorithm;

[0044] The rule module is used to monitor the abnormal status of the CRM system using a rule-based approach;

[0045] The fusion assessment module is used to conduct fusion assessment and early warning of abnormal status of the ERP system and CRM system;

[0046] The credibility verification module is used to verify the order data transmitted from the CRM system to the ERP system when the ERP system or CRM system enters the warning state, and to check the credibility of the data through one-way encryption;

[0047] The transmission load module is used to evaluate the transmission process of order data by using a logistic regression model and combining the load status of the ERP system and the CRM system;

[0048] The strategy optimization module is used to optimize the order management strategy based on the evaluation results of the logistic regression model.

[0049] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0050] This application conducts a comprehensive evaluation of the ERP system and the CRM system that provides data to the ERP system, monitors the stability of the ERP system using a method based on a machine learning algorithm, monitors the stability of the CRM system using a rule-based method, and verifies the credibility of data transmitted from the CRM system to the ERP system through a one-way encrypted verification method. It comprehensively controls the credibility of the transmission of original data between different systems, locates the abnormal failure links of the original data based on a logistic regression model, ensures the abnormal state of the data during transmission when the system fails, and provides optimization strategies based on the abnormal states of different systems to verify that the original order data is safe and reliable, and avoid data loss or distortion caused by the collapse of a single system or dual systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0052] Figure 1 The present invention is a flow chart of the method.

[0053] Figure 2 It is a system module diagram of the present invention. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] Example 1: Please refer to Figure 1 As shown, the present invention is an order management method based on ERP, and the specific steps include periodically checking the stability of the ERP system and the CRM system, monitoring the abnormal state of the ERP system by a method based on a machine learning algorithm, and monitoring the abnormal state of the CRM system by a rule-based method;

[0056] Evaluate and warn of abnormal status of ERP and CRM systems. When the ERP or CRM system enters the warning state, verify the order data transmitted from the CRM system to the ERP system and check the credibility of the data through one-way encryption.

[0057] Use logistic regression model to evaluate the order data transmission process in combination with the load status of ERP system and CRM system;

[0058] Based on the evaluation results of the logistic regression model, the order management strategy is optimized.

[0059] ERP system, or enterprise resource planning system, is an integrated information system that helps enterprises effectively manage and coordinate resources. It uses real-time data to integrate various departments and processes of the enterprise, such as finance, human resources, supply chain, production, sales, etc., to achieve unified data management and process automation.

[0060] CRM system, or customer relationship management system, is a technical solution used to manage the interaction between enterprises and existing and potential customers. It collects and stores various customer information, including contact information, purchase history, preferences and interaction records, and establishes automated sales processes to support marketing activities. The core goal of the CRM system is to improve customer satisfaction, loyalty and sales efficiency of the enterprise.

[0061] Use heartbeat detection to periodically check the stability of the ERP system and CRM system. Heartbeat detection is a mechanism for monitoring the health of the system. It confirms whether the system is operating normally by regularly sending and receiving signals. When a system fails or is interrupted, heartbeat detection can quickly detect it and trigger corresponding alarms or recovery operations. The signal sent and received regularly is the heartbeat.

[0062] Use heartbeat detection to periodically check the stability of the ERP system and CRM system. The method of periodic inspection is as follows:

[0063] Set up a system monitoring tool or framework for heartbeat detection, configure the monitored system components and set the frequency of sending heartbeat signals;

[0064] Write scripts to send heartbeat signals to the ERP system and CRM system regularly, receive the heartbeat signals through system monitoring tools and set thresholds;

[0065] If the heartbeat signal is not received within the set time, an early warning is triggered.

[0066] System monitoring tools or frameworks used for heartbeat detection include Nagios, Zabbix, Prometheus, etc. The heartbeat signals are in the form of HTTP requests, database queries, message queue messages, etc.

[0067] The abnormal status of the ERP system is monitored using a method based on machine learning algorithms. The specific method is as follows:

[0068] Collect log data, order records, operation records, and performance indicators of the ERP system;

[0069] Perform data preprocessing on the collected log data, order records, operation records, and performance indicators, remove or correct missing values, noise data, and outliers, and normalize the preprocessed data;

[0070] The time series features of normalized log data, order records, operation records, and performance indicators are used as the input of the feature vector. The time series data is divided into windows of fixed length using the sliding window method. The window numbers are marked as w = {1, 2, ... k}, where k is the number of windows and k is a positive integer, and the features of each window are extracted.

[0071] The extracted feature vector is input into the long short-term memory network model for training and the model output value is recorded. The model output value is marked as L op .

[0072] It should be noted that the characteristics of the window include mean, standard deviation, maximum value, minimum value, etc. The definition logic of the output value of the long short-term memory network model is that the larger the output value, the higher the probability of abnormality;

[0073] The LSTM network is a special type of recurrent neural network that can learn long-term dependencies in data. It controls the flow of information through the forget gate, input gate, and output gate. The three-gate structure allows the LSTM network to effectively maintain the flow of gradients during training, thereby learning long-term dependencies.

[0074] The method used for normalization is Euclidean distance normalization:

[0075] For each eigenvector x=[x1,x2,…,x n ], where n is the number of elements in the eigenvector, and n is a positive integer, there exists In the formula, v i is the normalized eigenvector, x i is the eigenvector before normalization.

[0076] The abnormal status of the CRM system is monitored through a rule-based method. The specific method is as follows:

[0077] Collect CRM system indicator extraction time data through system monitoring tools, including system log retrieval time, error rate calculation time, query response test time, and perform data preprocessing;

[0078] According to the window segmentation method of the sliding window, the system log retrieval time, error rate calculation time, and query response test time are marked and numbered, and the system log retrieval time is calibrated as S w The error rate calculation takes C w , the query response test takes T w , where w is the sequence number of the sliding window;

[0079] The rule model for configuring the system monitoring tool is: Where Ad is the rule response index, α and β are ln(S w +1) and The weight coefficient of , and α and β are both positive integers.

[0080] It should be noted that the system log retrieval time, error rate calculation time, and query response test time are all time series data corresponding to the windows divided by the sliding window in the machine learning algorithm model.

[0081] Conduct integrated assessment and early warning of abnormal status of ERP and CRM systems;

[0082] The classification threshold for the long short-term memory network model is preset to C t , the abnormal threshold of the rule model is preset as E t , when the calculated model output value L of the long short-term memory network model op Greater than or equal to the classification threshold C t , and the calculated rule response index Ad is greater than or equal to the abnormal threshold E t When the ERP system and the CRM system that transmits order data to the ERP system both experience abnormal failures, they are marked as level 1 abnormal warnings and send warning signals to management personnel;

[0083] When the model output value L of the long short-term memory network model is calculated op Less than the classification threshold C t , and the calculated rule response index Ad is greater than or equal to the abnormal threshold E t When the ERP system does not have any abnormal failure, but the CRM system that transmits order data to the ERP system has an abnormal failure, which is marked as a 2A level abnormal warning and sends a warning signal to the management personnel;

[0084] When the model output value L of the long short-term memory network model is calculated op Greater than or equal to the classification threshold C t , and the calculated rule response index Ad is less than the abnormal threshold E t When the ERP system has an abnormal failure, but the CRM system that transmits order data to the ERP system has no abnormal failure, it is marked as a 2B level abnormal warning and a warning signal is sent to the management personnel;

[0085] When the model output value L of the long short-term memory network model is calculated op Less than the classification threshold C t , and the calculated rule response index Ad is less than the abnormal threshold E t When , there is no abnormal failure in the ERP system and the CRM system that transmits order data to the ERP system, it is marked as normal status and the normal signal is recorded.

[0086] When the ERP system or CRM system enters the warning state, the order data transmitted from the CRM system to the ERP system is verified, and the credibility of the data is tested through one-way encryption;

[0087] When a level 1 abnormal warning, a level 2A abnormal warning, or a level 2B abnormal warning is marked, the order data transmitted from the CRM system to the ERP system is verified, and the method for verifying the credibility of the data through one-way encryption is as follows:

[0088] The CRM system converts the benchmark order data into a string or binary format and sends it to the ERP system. The CRM system also uses a hash algorithm to calculate a benchmark hash value based on the string or binary format of the benchmark order data. The benchmark hash value is then saved by the ERP system.

[0089] When the ERP system receives the order data from the CRM system, it calculates the hash value of the order data according to the hash algorithm and verifies the calculated hash value with the reference hash value. When the verification is consistent, the data verification is marked as passed and the verification value C is set. v When the verification is inconsistent, the data verification is marked as failed and the verification value C is set. v is 0.

[0090] Use logistic regression model to evaluate the order data transmission process in combination with the load status of ERP system and CRM system;

[0091] The calibrated resource load rate is L r , where r is the number of servers with different loads, and r = 1, 2, ... u}, where u is the number of load servers, and u is a positive integer, and the component load rate is T p , where p is the number of different computing power components in the same load server, and p = 1, 2, ... q}, where q is the number of computing power components in the load server, and q is a positive integer. A logistic regression model is established to calculate the transmission load index T rs , the calculation expression is

[0092] It should be noted that the computing power components in the load server include CPU, memory, disk, network traffic, and I / O interface.

[0093] Optimize order management strategies based on the evaluation results of the logistic regression model;

[0094] Preset the transmission critical threshold T for the logistic regression model al , when the manager receives a Level 1 abnormality warning and the calculated transmission load index T rs Greater than the preset transmission critical threshold T al When the ERP system, CRM system and order data input are abnormal, the order data is unreliable, and the notification optimization strategy is to restart the ERP system and CRM system and manually input the order data;

[0095] When the manager receives a Level 1 abnormality warning and the calculated transmission load index T rs Less than or equal to the preset transmission critical threshold T al When the ERP system and CRM system are abnormal, the output of the order data is abnormal, and there is a risk of channel distortion of the order data. The notification optimization strategy is to restart the ERP system and CRM system and repeat the verification of the order data;

[0096] When the manager receives a Level 2A abnormal warning and the calculated transmission load index T rs Greater than the preset transmission critical threshold T al When the CRM system and order data input are abnormal, there is a risk of source distortion of order data. The notification optimization strategy is to restart the CRM system and manually check the order source information;

[0097] When the manager receives a Level 2A abnormal warning and the calculated transmission load index T rs Less than or equal to the preset transmission critical threshold T al When the output of the CRM system and order data are abnormal, there is a risk of channel distortion of the order data. The notification optimization strategy is to restart the CRM system and repeat the verification of the order data;

[0098] When the manager receives a Level 2B abnormality warning and the calculated transmission load index T rs Greater than the preset transmission critical threshold T al When the ERP system and order data input are abnormal, there is a risk of source distortion of order data. The notification optimization strategy is to restart the ERP system and manually check the order source information;

[0099] When the manager receives a Level 2B abnormality warning and the calculated transmission load index T rs Less than or equal to the preset transmission critical threshold T al When , the output of the ERP system and order data are abnormal, and there is a risk of channel distortion of the order data. The notification optimization strategy is to restart the ERP system and repeat the verification of the order data.

[0100] Example 2: Please refer to Figure 2 As shown, the ERP-based order management system includes a periodic verification module, a machine learning module, a rule module, a fusion evaluation module, a credibility verification module, a transmission load module, and a strategy optimization module;

[0101] The periodic verification module is used to periodically check the stability of the ERP system and CRM system;

[0102] The machine learning module is used to monitor the abnormal status of the ERP system according to a method based on a machine learning algorithm;

[0103] The rule module is used to monitor the abnormal status of the CRM system using a rule-based approach;

[0104] The fusion assessment module is used to conduct fusion assessment and early warning of abnormal status of the ERP system and CRM system;

[0105] The credibility verification module is used to verify the order data transmitted from the CRM system to the ERP system when the ERP system or CRM system enters the warning state, and to check the credibility of the data through one-way encryption;

[0106] The transmission load module is used to evaluate the transmission process of order data by using a logistic regression model and combining the load status of the ERP system and the CRM system;

[0107] The strategy optimization module is used to optimize the order management strategy based on the evaluation results of the logistic regression model.

[0108] This application conducts a comprehensive evaluation of the ERP system and the CRM system that provides data to the ERP system, monitors the stability of the ERP system using a method based on a machine learning algorithm, monitors the stability of the CRM system using a rule-based method, and verifies the credibility of data transmitted from the CRM system to the ERP system through a one-way encrypted verification method. It comprehensively controls the credibility of the transmission of original data between different systems, locates the abnormal failure links of the original data based on a logistic regression model, ensures the abnormal state of the data during transmission when the system fails, and provides optimization strategies based on the abnormal states of different systems to verify that the original order data is safe and reliable, and avoid data loss or distortion caused by the collapse of a single system or dual systems.

Claims

1. The order management method based on ERP is characterized by: The specific steps include periodic inspection of the stability of the ERP system and the CRM system, monitoring the abnormal status of the ERP system using a method based on machine learning algorithms, and monitoring the abnormal status of the CRM system using a rule-based method; Evaluate and warn of abnormal status of ERP and CRM systems. When the ERP or CRM system enters the warning state, verify the order data transmitted from the CRM system to the ERP system and check the credibility of the data through one-way encryption. Use logistic regression model to evaluate the order data transmission process in combination with the load status of ERP system and CRM system; Based on the evaluation results of the logistic regression model, the order management strategy is optimized.

2. The ERP-based order management method according to claim 1, characterized in that: The method for periodically checking the stability of ERP and CRM systems is: Set up a system monitoring tool or framework for heartbeat detection, configure the monitored system components and set the frequency of sending heartbeat signals; Write scripts to send heartbeat signals to the ERP system and CRM system regularly, receive the heartbeat signals through system monitoring tools and set thresholds; If the heartbeat signal is not received within the set time, an early warning is triggered.

3. The ERP-based order management method according to claim 1, characterized in that: The abnormal status of the ERP system is monitored using a method based on machine learning algorithms. The specific method is as follows: Collect log data, order records, operation records, and performance indicators of the ERP system; Perform data preprocessing on the collected log data, order records, operation records, and performance indicators, remove or correct missing values, noise data, and outliers, and normalize the preprocessed data; The time series features of normalized log data, order records, operation records, and performance indicators are used as the input of the feature vector. The time series data is divided into windows of fixed length using the sliding window method. The window numbers are marked as w = {1, 2, ... k}, where k is the number of windows and k is a positive integer, and the features of each window are extracted. The extracted feature vector is input into the long short-term memory network model for training and the model output value is recorded. The model output value is marked as L op .

4. The ERP-based order management method according to claim 3, characterized in that: The method used for normalization is Euclidean distance normalization: For each eigenvector x=[x1,x2,…,x n ], where n is the number of elements in the eigenvector, and n is a positive integer, there exists In the formula, v i is the normalized eigenvector, x i is the eigenvector before normalization.

5. The ERP-based order management method according to claim 3, characterized in that: The abnormal status of the CRM system is monitored through a rule-based method. The specific method is as follows: Collect CRM system indicator extraction time data through system monitoring tools, including system log retrieval time, error rate calculation time, query response test time, and perform data preprocessing; According to the window segmentation method of the sliding window, the system log retrieval time, error rate calculation time, and query response test time are marked and numbered, and the system log retrieval time is calibrated as S w The error rate calculation takes C w , the query response test takes T w , where w is the sequence number of the sliding window; The rule model for configuring the system monitoring tool is: Where Ad is the rule response index, α and β are ln(S w +1) and The weight coefficient of , and α and β are both positive integers.

6. The ERP-based order management method according to claim 5, characterized in that: The logic for integrated evaluation and early warning of abnormal status of ERP and CRM systems is as follows: The classification threshold for the long short-term memory network model is preset to C t , the abnormal threshold of the rule model is preset as E t , when the calculated model output value L of the long short-term memory network model op Greater than or equal to the classification threshold C t , and the calculated rule response index Ad is greater than or equal to the abnormal threshold E t When the ERP system and the CRM system that transmits order data to the ERP system both experience abnormal failures, they are marked as level 1 abnormal warnings and send warning signals to management personnel; When the model output value L of the long short-term memory network model is calculated op Less than the classification threshold C t , and the calculated rule response index Ad is greater than or equal to the abnormal threshold E t When the ERP system does not have any abnormal failure, but the CRM system that transmits order data to the ERP system has an abnormal failure, which is marked as a 2A level abnormal warning and sends a warning signal to the management personnel; When the model output value L of the long short-term memory network model is calculated op Greater than or equal to the classification threshold C t , and the calculated rule response index Ad is less than the abnormal threshold E t When the ERP system has an abnormal failure, but the CRM system that transmits order data to the ERP system has no abnormal failure, it is marked as a 2B level abnormal warning and a warning signal is sent to the management personnel; When the model output value L of the long short-term memory network model is calculated op Less than the classification threshold C t , and the calculated rule response index Ad is less than the abnormal threshold E t When , there is no abnormal failure in the ERP system and the CRM system that transmits order data to the ERP system, it is marked as normal status and the normal signal is recorded.

7. The ERP-based order management method according to claim 1, characterized in that: The verification method for verifying the credibility of data through one-way encryption is: The CRM system converts the benchmark order data into a string or binary format and sends it to the ERP system. The CRM system also uses a hash algorithm to calculate a benchmark hash value based on the string or binary format of the benchmark order data. The benchmark hash value is then saved by the ERP system. When the ERP system receives the order data from the CRM system, it calculates the hash value of the order data according to the hash algorithm and verifies the calculated hash value with the reference hash value. When the verification is consistent, the data verification is marked as passed and the verification value C is set. v When the verification is inconsistent, the data verification is marked as failed and the verification value C is set. v is 0.

8. The ERP-based order management method according to claim 7, characterized in that: The logic of using the logistic regression model to evaluate the order data transmission process combined with the load status of the ERP system and the CRM system is as follows; The calibrated resource load rate is L r , where r is the number of servers with different loads, and r = 1, 2, ... u}, where u is the number of load servers, and u is a positive integer, and the component load rate is T p , where p is the number of different computing power components in the same load server, and p = 1, 2, ... q}, where q is the number of computing power components in the load server, and q is a positive integer. A logistic regression model is established to calculate the transmission load index T rs , the calculation expression is 9. The ERP-based order management method according to claim 8, characterized in that: Optimize order management strategies based on the evaluation results of the logistic regression model; Preset the transmission critical threshold T for the logistic regression model al , when the manager receives a Level 1 abnormality warning and the calculated transmission load index T rs Greater than the preset transmission critical threshold T al When the ERP system, CRM system and order data input are abnormal, the order data is unreliable, and the notification optimization strategy is to restart the ERP system and CRM system and manually input the order data; When the manager receives a Level 1 abnormality warning and the calculated transmission load index T rs Less than or equal to the preset transmission critical threshold T al When the ERP system and CRM system are abnormal, the output of the order data is abnormal, and there is a risk of channel distortion of the order data. The notification optimization strategy is to restart the ERP system and CRM system and repeat the verification of the order data; When the manager receives a Level 2A abnormal warning and the calculated transmission load index T rs Greater than the preset transmission critical threshold T al When the CRM system and order data input are abnormal, there is a risk of source distortion of order data. The notification optimization strategy is to restart the CRM system and manually check the order source information; When the manager receives a Level 2A abnormal warning and the calculated transmission load index T rs Less than or equal to the preset transmission critical threshold T al When the output of the CRM system and order data are abnormal, there is a risk of channel distortion of the order data. The notification optimization strategy is to restart the CRM system and repeat the verification of the order data; When the manager receives a Level 2B abnormality warning and the calculated transmission load index T rs Greater than the preset transmission critical threshold T al When the ERP system and order data input are abnormal, there is a risk of source distortion of order data. The notification optimization strategy is to restart the ERP system and manually check the order source information; When the manager receives a Level 2B abnormality warning and the calculated transmission load index T rs Less than or equal to the preset transmission critical threshold T al When , the output of the ERP system and order data are abnormal, and there is a risk of channel distortion of the order data. The notification optimization strategy is to restart the ERP system and repeat the verification of the order data.

10. An ERP-based order management system, used to implement the management method according to any one of claims 1 to 9, characterized in that: It includes periodic verification module, machine learning module, rule module, fusion evaluation module, credibility verification module, transmission load module and strategy optimization module; The periodic verification module is used to periodically check the stability of the ERP system and CRM system; The machine learning module is used to monitor the abnormal status of the ERP system according to a method based on a machine learning algorithm; The rule module is used to monitor the abnormal status of the CRM system using a rule-based approach; The fusion assessment module is used to conduct fusion assessment and early warning of abnormal status of the ERP system and CRM system; The credibility verification module is used to verify the order data transmitted from the CRM system to the ERP system when the ERP system or CRM system enters the warning state, and to check the credibility of the data through one-way encryption; The transmission load module is used to evaluate the transmission process of order data by using a logistic regression model and combining the load status of the ERP system and the CRM system; The strategy optimization module is used to optimize the order management strategy based on the evaluation results of the logistic regression model.

Citation Information

Patent Citations

  • Customer information management system

    CN106529756A

  • Digital value monitoring system and digital value evaluation method

    CN115471060A

  • Distributed energy storage informatization control and operation platform architecture

    CN115827579A

  • Order management system based on block chain technology

    CN117689442A

  • Communication information security risk early warning management and control method and system based on big data

    CN117955712A

Cited By

  • Intelligent order generation and management system

    CN120806870A