Accurate order forecasting system and forecasting method thereof

Abstract

This invention discloses an accurate order prediction system and method. The accurate order prediction system includes an input device, a storage device, a processing device, and an output device. The input device includes an order input interface to receive orders input by customers. The storage device stores orders, historical order information, and multiple algorithms. The processing device accesses the storage device to execute the multiple algorithms, establishes an accurate order prediction model to obtain accurate order information, and the accurate order prediction model includes an order evaluation module, an order placement prediction module, and an order correction module. The output device is connected to the processing device and the storage device to output the accurate order information.

Description

Technical Field

[0001] The present invention relates to a precise order prediction system and method, and more particularly to a precise order prediction system and method that accurately predicts order status using the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) and artificial intelligence (AI) algorithms. Prior Art

[0002] Before manufacturing any product, sales personnel receive prospective customer orders and plan production based on the required quantity, hoping to fulfill the order by the customer's deadline. However, from the time the order is placed to actual production, and even to delivery and sales, various factors, including industry, environmental, and market trends, can affect the quantity and delivery date of the product. Customers canceling or adjusting orders is common, and when manufacturers accept orders and plan production, these adjustments or even cancellations often disrupt the entire production line, increasing unnecessary production costs or creating unsold inventory.

[0003] Currently, production scheduling can adjust production capacity based on changes in order demand. However, by the time order changes occur, the associated manufacturing costs have already been invested, and even with schedule adjustments, costs may still be incurred. If the availability of orders could be predicted beforehand and used to plan production schedules, this would help reduce unnecessary costs and waste, thereby enhancing manufacturers' competitiveness.

[0004] In light of this, existing order management systems struggle to accurately predict the status of valid orders and are unable to provide the correct information needed for production scheduling. To address this, the inventors of this invention have devised a precise order forecasting system and method to address these shortcomings and enhance their practical application in the industry. Summary of the Invention

[0005] In view of the above-mentioned problems of conventional technology, the purpose of the present invention is to provide an accurate order prediction system and prediction method thereof to solve the problem of errors generated in the actual operation of conventional order management systems.

[0006] According to one objective of the present invention, a precise order prediction system is provided, comprising an input device, a storage device, a processing device, and an output device. The input device includes an order input interface for receiving orders input by orderers, the orders including order quantities and order information. A storage device is connected to the input device to receive orders and stores orders, historical order information, and a plurality of algorithms. A processing device is connected to the storage device and accesses the storage device to execute a plurality of algorithms to establish a precise order prediction model. The precise order prediction model comprises an order evaluation module, an order destination prediction module, and an order correction module. The order evaluation module inputs the order quantity and order information into a production planning program to generate a production delivery schedule and estimate initial order information. The order destination prediction module uses historical order information to perform a destination analysis of the initial order information, assessing the effectiveness of executed orders and identifying invalid orders whose effectiveness falls below a preset threshold. The order correction module corrects the initial order information by eliminating invalid orders to generate precise order information. The output device is connected to the processing device and the storage device to output precise order information.

[0007] Preferably, the landing point analysis process may include a Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) and an artificial intelligence (AI) algorithm.

[0008] Preferably, the order information may include the ordering manufacturer, order priority, delivery date and payment information.

[0009] Preferably, the order history information may include completed order information and canceled order information.

[0010] Preferably, the output device can output warning information of invalid orders.

[0011] According to one purpose of the present invention, a method for accurately predicting orders is proposed, which is applicable to an accurate order prediction system including an input device, a storage device, a processing device and an output device. The method for accurately predicting orders includes the following steps: setting an order input interface of the input device, receiving an order input by an orderer through the order input interface, the order including an order quantity and order information; storing the order in a storage device, and storing historical order information and a plurality of algorithms in the storage device; accessing the storage device through a processor, executing a plurality of algorithms, and establishing an accurate order prediction model, the accurate order prediction model including an order evaluation module, an order landing prediction module and an order correction module; inputting the order quantity and order information into a production planning program through the order evaluation module to form a production delivery schedule, estimating initial order information, performing a landing analysis program of the initial order information based on the historical order information through the order landing prediction module, evaluating the effectiveness of the executed order, evaluating invalid orders whose effectiveness is lower than a preset threshold, eliminating invalid orders through the order correction module to correct the initial order information and form accurate order information; and outputting the accurate order information through the output device.

[0012] Preferably, the landing point analysis process may include a Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) and an artificial intelligence (AI) algorithm.

[0013] Preferably, the order information may include the ordering manufacturer, order priority, delivery date and payment information.

[0014] Preferably, the order history information may include completed order information and canceled order information.

[0015] Preferably, the output device can output warning information of the invalid order.

[0016] As described above, the precise order forecasting system and forecasting method according to the present invention may have one or more of the following advantages:

[0017] (1) This order precision forecasting system and its forecasting method can obtain accurate order information through the order precision forecasting model, eliminate invalid orders to reduce unnecessary manufacturing costs, and enhance the manufacturer's production competitiveness.

[0018] (2) This precise order prediction system and its prediction method can predict the order placement through the ideal solution similarity order preference method of the placement analysis program to obtain quantitative information on the effectiveness of the order, and then cooperate with artificial intelligence algorithms to improve the accuracy of judgment and improve the accuracy of the overall judgment.

[0019] (3) This order prediction system and its prediction method provide accurate order information through the output device to adjust the production schedule, and prompt invalid orders through warning information to assist order management personnel in order control and adjustment. Simple diagram description

[0020] To make the technical features, contents, advantages and effects of the present invention more apparent, the present invention is described in detail in the form of embodiments with reference to the accompanying drawings as follows: FIG1 is a block diagram of an accurate order forecasting system according to an embodiment of the present invention. FIG2 is a schematic diagram of the landing point analysis process according to an embodiment of the present invention. FIG3 is a flow chart of a method for accurate order prediction according to an embodiment of the present invention. Implementation Method

[0021] To facilitate understanding of the technical features, contents, advantages, and effects achievable by the present invention, the present invention is described in detail below in the form of embodiments, with reference to the accompanying drawings. The figures used therein are for illustration and auxiliary description purposes only and may not reflect the actual proportions and precise configurations of the present invention after implementation. Therefore, the proportions and configurations in the attached figures should not be interpreted to limit the scope of the present invention in actual implementation. This is to be noted in advance.

[0022] Please refer to Figure 1, which is a block diagram of an accurate order forecasting system according to an embodiment of the present invention. As shown, accurate order forecasting system 10 includes an input device 11, a storage device 12, a processing device 13, and an output device 14. Input device 11 includes an order input interface 111 for receiving orders 121 input by subscribers. Input device 11 can be a physical computer device, with order input interface 111 designed on the computer device's display input interface, allowing subscribers to enter orders 121 through the interface. In other embodiments, input device 11 can also be a virtual network platform, with an order input webpage set up via the internet as order input interface 111, allowing subscribers to enter orders 121 online. Order 121 can include the order quantity and order information, where the order information can include the ordering vendor, order priority, delivery date, and payment information.

[0023] Storage device 12 is connected to input device 11 to receive orders 121. Storage device 12 stores orders 121, historical order information 122, and a plurality of algorithms 123. Storage device 12 can be a computer device's read-only memory, flash memory, disk, or other storage media, or a database on a server or cloud device. When order 121 is received from input device 11, it is stored in storage device 12. The subsequent predictive analysis algorithm is also stored in storage device 12. Processing device 13 is connected to storage device 12. Processing device 13 includes a central processing unit (CPU), microprocessor, graphics processor, etc., found in a computer or server device. Processing device 13 executes various control commands to access storage device 12 and establishes an accurate order prediction model 131 by executing the analysis programs of each algorithm 123.

[0024] The order accuracy forecasting model 131 includes an order assessment module 132, an order destination forecasting module 133, and an order correction module 134. The order assessment module 132 inputs order quantities and order information into the production planning process to generate a production delivery schedule and estimate initial order information. Based on the order information and order quantities for each product, a production schedule is generated for each product. Manufacturing and delivery dates are planned based on quantity and lead time. When the production delivery schedule exceeds production capacity, order delivery deadlines must be adjusted or production capacity increased. If orders 121 contain duplicates, unreasonable orders, or potentially cancelable items, this will affect the production schedule. Therefore, predictive analysis is intended to identify these problematic orders.

[0025] The order destination prediction module 133 uses historical order information to perform a destination analysis of initial order information, evaluating the effectiveness of executed orders and identifying invalid orders whose effectiveness falls below a preset threshold. Since the storage device 12 stores historical order information 122, including completed and canceled order information, the completion status of historical orders can be obtained from historical order information 122, and this historical information serves as the basis for initial order evaluation. Please refer to Figure 2, which is a schematic diagram of the destination analysis process according to an embodiment of the present invention. The destination analysis process may include the Top-of-Preferences-by-Ideal-Solutions Similarity (TOPSIS) method 211 and an artificial intelligence (AI) algorithm 212. Ideal solution similarity order preference method 211 is a multi-attribute decision-making algorithm. The idealized goals during planning include positive and negative ideal goals, namely, the best and worst solutions. Initial order information 201 is analyzed using similarity order preference method 211 to determine the distance between each order and the best and worst targets, respectively. The order's quality is determined by comparing its landing points. If the order information is greater than a predetermined distance from the best target or less than a predetermined distance from the worst target, the order is deemed to have a cancellation or unfulfilled rate exceeding a certain threshold and is deemed invalid 202. Artificial intelligence algorithm 212 utilizes various deep learning or machine learning algorithms to improve the accuracy and efficiency of this landing point analysis.

[0026] Algorithms can include data searches. After entering relevant keywords, the algorithm uses the keywords to locate records within a database field, thereby retrieving related records, attributes, or related data. Searches can include searches within a sequential structure or searches utilizing an index structure. Sequential searches involve sequentially arranging data one record at a time or symbol at a time, without any other assistance. Sequential structures can be categorized as either unsorted or sorted based on key columns. Unsorted sequential structures require no data organization and are therefore suitable for frequently changing data. Conversely, for data that rarely changes, sorting the data and then using binary search or interpolation search methods can improve search speed.

[0027] Searching using an index structure is an additional tool to aid in searching. These include direct indexing, tree-structured indexing, or the use of hash tables. Hash tables involve designing hash functions to convert data keys into storage locations. In real-world situations, data collisions can occur, where different keys are converted to the same location. Therefore, hash function design must balance collision reduction with computational simplicity. Common hash functions include division, squaring, folding, and bitwise analysis.

[0028] The order placement prediction module 133 identifies predicted invalid orders 202. The order correction module 134 then corrects the initial order information 201 by eliminating invalid orders 202 to generate accurate order information 141. This eliminates duplicate, potentially canceled, or uncompleted order information, re-forming accurate order information 141. The output device 14 is connected to the processing device 13 and the storage device 12. The accurate order information 141 obtained by the accurate order prediction model 131 can be output via the output device 14. The output device 14 can be the same computer device as the input device 11, presenting the analysis results on a display interface or transmitting the relevant information via a data transmission device to the computer of an order planner or supervisor, who can then reschedule the production of the relevant products based on the accurate order information 141.

[0029] After the production and delivery of the goods are completed, the information of the completed order can be stored in the storage device 12 as new historical order information 122. For the invalid orders 202 that are eliminated, warning information can be sent to the relevant processing personnel through the output device 14.

[0030] Please refer to Figure 3, which is a flow chart of a precise order prediction method according to an embodiment of the present invention. This precise order prediction method can be applied to a precise order prediction system comprising an input device, a storage device, a processing device, and an output device as described in the aforementioned embodiment. In this precise order prediction system, the precise order prediction method includes the following steps (S1-S5):

[0031] Step S1: An order input interface is provided on the input device to receive orders from subscribers via the order input interface. Orders include order quantities and order information. The order input interface is provided on the display interface of the input device or on a webpage on a network platform, allowing subscribers to enter orders through the interface. Orders may include order quantities and order information, including the ordering vendor, order priority, delivery date, and payment information.

[0032] Step S2: Storing the order in a storage device, and storing historical order information and multiple algorithms in the storage device. The order input in the above steps can be stored in a storage device such as a computer device's storage media or a cloud database. The storage device can also store historical order data and subsequent predictive analysis algorithms. The historical order data includes completed order information and canceled order information.

[0033] Step S3: The processor accesses the storage device and executes multiple algorithms to establish an accurate order prediction model. The accurate order prediction model includes an order assessment module, an order destination prediction module, and an order correction module. The processing device includes a central processing unit, a microprocessor, an image processing processor, etc. in a computer or server device. The processing device can be connected to the processing device and execute various control instructions to access the storage device. The accurate order prediction model is established by executing various algorithm programs in the storage device.

[0034] Step S4: The order evaluation module inputs the order quantity and order information into the production planning program to generate a production delivery schedule and estimate initial order information. The order destination prediction module then uses historical order information to perform a destination analysis of the initial order information, assessing the effectiveness of executed orders. Invalid orders whose effectiveness falls below a preset threshold are evaluated, and the order correction module eliminates invalid orders to correct the initial order information, generating accurate order information. In the accurate order prediction model, the order evaluation module inputs the order quantity and order information into the production planning program to perform production scheduling and estimate manufacturing and delivery dates as the initial order information. The initial order information may include duplicate, invalid, or potentially canceled orders. Therefore, the order destination prediction module further evaluates the destination analysis of the initial order information to identify related invalid orders. The destination analysis process may include an ideal solution similarity order preference method and an artificial intelligence algorithm. Using historical order data, the distance between the initial order and the optimal and worst targets is calculated. These distances are then compared with preset thresholds to identify invalid orders. The order correction module eliminates the predicted invalid orders, resulting in accurate order information.

[0035] Step S5: Output the accurate order information via the output device. The accurate order information obtained in the above steps is output via the output device to order planners or managers, who can then replan the product production schedule to achieve optimal order efficiency. In other embodiments, the output device can send warnings regarding invalid orders to the corresponding processing personnel, allowing them to confirm the order or place a new one. Furthermore, after the production and delivery of the product is completed, the completed order information can be stored in the storage device as new historical order information.

[0036] The above description is for illustrative purposes only and is not intended to be limiting. Any equivalent modifications or variations that do not depart from the spirit and scope of the present invention should be included in the scope of the appended patent applications.

[0037] 10: Order Accurate Forecasting System 11: Input device 12: Storage device 13: Processing device 14: Output device 111: Input Interface 121:Order 122: Historical order information 123: Multiple Algorithms 131: Order Accurate Prediction Model 132: Order Evaluation Module 133: Order location prediction module 134: Order correction module 141: Accurate order information 201: Initial order information 202: Invalid order 211: Ideal Solution Similarity Order Preference Method 212: Artificial Intelligence Algorithms S1~S5: Steps

Claims

1. An order prediction system, comprising: an input device including an order input interface for receiving an order input by a customer, the order including an order quantity and order information, the order information including the ordering manufacturer, order priority, delivery date, and payment information; a storage device connected to the input device for receiving the order, the storage device storing the order, historical order information, and a plurality of algorithms, wherein the historical order information includes completed order information and cancelled order information; and a processing device connected to the storage device for accessing the storage device to execute the plurality of algorithms to establish an order prediction model, the order prediction model including the following modules: an order evaluation module for inputting the order quantity and order information into a production planning program to form a production delivery schedule and predicting initial order information based on capacity load; An order placement prediction module performs an ideal solution similarity order preference method on the initial order information based on historical order information. It sets an optimal target and a worst target, and calculates the distance between the initial order information and the optimal target and the worst target, respectively. If the distance to the optimal target is greater than a predetermined distance or the distance to the worst target is less than a predetermined distance, the order is determined to be an invalid order. An order correction module corrects the initial order information by removing invalid orders to form accurate order information. An output device is connected to the processing device and the storage device to output the accurate order information.

2. The order prediction system as described in claim 1, wherein the output device outputs a warning message for the invalid order.

3. An order prediction method applicable to an order prediction system comprising an input device, a storage device, a processing device, and an output device, the order prediction method comprising the following steps: setting an order input interface on the input device, receiving an order input by a customer through the order input interface, the order including an order quantity and order information, the order information including the ordering manufacturer, order priority, delivery date, and payment information; storing the order in the storage device, and storing historical order information and a plurality of algorithms in the storage device, the historical order information including completed order information and cancelled order information; accessing the storage device through the processing device, executing the plurality of algorithms, and establishing an order prediction model, the order prediction model including an order evaluation module, an order placement prediction module, and an order correction module; The order evaluation module inputs the order quantity and order information into a production planning program to generate a production delivery schedule. Based on the capacity load, an initial order information is estimated. The order placement prediction module uses historical order information to perform a placement analysis program based on the ideal solution similarity order preference method on the initial order information. An optimal target and a worst-case target are set, and the distances between the initial order information and the optimal and worst-case targets are calculated respectively. If the distance to the optimal target is greater than a predetermined distance or the distance to the worst-case target is less than a predetermined distance, the order is determined to be an invalid order. The order correction module removes the invalid order to correct the initial order information, generating accurate order information. The accurate order information is output through the output device.

4. The order prediction method as described in claim 3, wherein the output device outputs a warning message for the invalid order.

Images