Energy-saving control system and method for laser powder bed fusion production process

Abstract

This invention discloses an energy-saving control system and method for a laser powder bed melting production process, belonging to the fields of green manufacturing and additive manufacturing technology. The system includes: a physical sensing layer containing a first RFID tag mounted on a printer substrate, RFID readers / writers located at each production device, and digital meters; a data processing layer for constructing a dynamic energy consumption matrix with parts as rows and processes as columns; and a control execution layer for generating and executing adjustment instructions for subsequent batch production plans based on energy efficiency calculation results. The method includes: triggering the meters to collect the total energy consumption of each process through RFID read events; allocating the total energy consumption of each process to each part and filling it into the dynamic energy consumption matrix; comparing the calculated energy consumption value with historical benchmark values ​​to calibrate the energy efficiency level, and automatically applying improvement measures corresponding to low energy efficiency to the next batch. This invention achieves individual energy consumption traceability and batch-to-batch closed-loop optimization throughout the laser powder bed melting process, effectively reducing overall energy consumption.

Description

Technical Field

[0001] This invention belongs to the field of green manufacturing and additive manufacturing technology, and specifically relates to an energy-saving control system and method for laser powder bed melting production process. Background Technology

[0002] Laser powder bed fusion (LPBF) is a mainstream technology in metal additive manufacturing and has been widely used in aerospace, automotive, and medical fields. Unlike the subtractive manufacturing production model of "processing individual blanks sequentially," the typical production model of LPBF involves printing multiple parts simultaneously on the same substrate, then transferring the entire board to an electrical discharge wire cutting machine to separate individual parts, and finally transporting them via an automated guided vehicle (AGV) to a CNC machine tool for finishing. This unique production process brings several challenges that existing energy consumption control technologies struggle to address.

[0003] Currently, there are several patents that use radio frequency identification (RFID) technology for energy consumption monitoring in manufacturing processes.

[0004] Chinese patent CN104076768B discloses a "method for acquiring and controlling energy-saving power consumption during order execution," which uses digital meters and RFID to acquire the real-time power consumption of each material at each workstation during order execution and constructs an power consumption matrix. This method primarily targets the order hierarchy in discrete manufacturing (order → sub-process → production task → workstation), and its energy consumption aggregation is based on a one-to-one relationship between workstation and material, assuming that the material already exists before processing and can be tagged with RFID. However, before LPBF printing, parts exist in powder form, without physical parts, making pre-labeling impossible; furthermore, this method does not involve the allocation of total batch energy consumption and cannot handle the special problem of individual energy consumption not being directly measurable when multiple parts are printed on the same board.

[0005] Chinese patent CN111414983B discloses "A method for controlling power consumption in a machining workshop based on radio frequency identification and scheduling." This method, in a machining workshop, uses RFID read events to trigger dynamic updates to the workshop's work schedule to address energy consumption anomalies during processing. This method targets single-piece processing in subtractive manufacturing, and its core lies in adjusting the processing sequence (scheduling scheme) based on real-time energy consumption feedback to reduce the overall energy consumption of the workshop. However, this method also relies on the premise that "the part already has an RFID tag attached before processing" and does not address the technical issues unique to additive manufacturing, such as multi-part printing on the same board, batch energy consumption allocation, and build direction optimization. Furthermore, this method only focuses on the internal workings of the machining workshop and does not cover the multi-process coupled energy consumption of the entire LPBF process (LPBF printing, wire EDM, AGV transportation, and CNC finishing).

[0006] In summary, existing patents all target subtractive manufacturing or general discrete manufacturing, and their technical solutions cannot be directly transferred to the LPBF production process. Specifically:

[0007] (1) There is no physical part before LPBF printing, so it is impossible to attach RFID tags in advance. Existing methods require labeling to be completed before processing, which contradicts the powder → physical forming process of LPBF.

[0008] (2) Printing multiple parts on the same board makes it impossible to directly measure the energy consumption of individual printing. Existing methods collect energy consumption based on the one-to-one relationship between workstation and material, which cannot handle the scenario of "printing a batch of parts and sharing the total energy consumption of multiple parts".

[0009] (3) The energy consumption of LPBF is affected by special variables such as the construction direction and the volume of the supporting structure, which do not exist in subtractive manufacturing. The energy consumption model of the existing method does not include the above variables.

[0010] (4) The entire LPBF process involves multiple processes such as printing, wire cutting, AGV transportation, and CNC precision machining. There is a coupling relationship between the energy consumption of the processes (for example, the construction direction and support volume in the printing stage will affect the contact area and processing difficulty of subsequent wire cutting). Existing methods only focus on the energy consumption control of a single process or within the workshop, and lack cross-process energy consumption matrix and closed-loop control mechanism. Summary of the Invention

[0011] This invention aims to solve the technical problems in existing technologies, such as the inability to trace individual energy consumption in LPBF (Laser-Based Powder Bed Fusion) multi-part simultaneous printing scenarios, the inability to reasonably allocate total batch energy consumption, the lack of real-time monitoring and energy efficiency calculation for multi-process coupling throughout the entire process, and the inability to perform closed-loop optimization between batches based on energy efficiency feedback. It provides an energy-saving control system and method for laser powder bed fusion production processes.

[0012] In a first aspect, the present invention provides an energy-saving control system for a laser powder bed melting production process, comprising:

[0013] The physical sensing layer includes a first radio frequency identification tag disposed on the substrate of the laser powder bed fusion printer, radio frequency identification readers and writers disposed at each production equipment, and a digital meter. Each production equipment includes at least a laser powder bed fusion printer, an electrical discharge wire cutting machine, an automated guided vehicle, and a CNC machine tool.

[0014] The data processing layer, which is communicatively connected to the physical sensing layer, is used to construct a dynamic energy consumption matrix with parts as rows and processes as columns. The processes include at least laser powder bed fusion printing, electrical discharge wire cutting, automated guided vehicle transportation, and CNC precision machining.

[0015] The control execution layer, which is communicatively connected to the data processing layer, is used to generate and execute adjustment instructions for subsequent batch production plans based on the energy efficiency calculation results of each part or process in the dynamic energy consumption matrix.

[0016] Preferably, the first RFID tag is a high-temperature resistant metal-based tag that stores the batch number corresponding to the substrate; the system also includes a second RFID tag, which is attached to each part after cutting and separation, and stores the part number of the part, which is used to trigger an RFID read event at the corresponding equipment during each process flow, so as to trigger the digital meter to collect and record energy consumption data.

[0017] Preferably, the adjustment instructions generated by the control execution layer include at least one of optimizing the part construction direction, printing process parameters, part batching scheme, or cutting parameters.

[0018] An energy-saving control method for a laser powder bed melting production process, using the above-mentioned system, includes the following steps:

[0019] The radio frequency identification (RFID) read event in the physical sensing layer triggers the digital meter of the corresponding process to collect and record the total energy consumption of that process.

[0020] Using the data processing layer, the total energy consumption of each process is allocated to each part and filled into the corresponding position of the dynamic energy consumption matrix. Based on the dynamic energy consumption matrix, the energy consumption value of each part in each process and the entire process is calculated.

[0021] Using the control execution layer, the calculated energy consumption value is compared with the historical benchmark value. Based on the comparison result, the energy efficiency level is calibrated, and the improvement measures corresponding to the parts or processes with energy efficiency levels lower than the preset level are automatically applied to the production plan of the next batch.

[0022] Preferably, the printing process further includes an interlayer energy consumption monitoring step: after each layer of powder melting is completed, the energy consumption per unit volume of that layer is calculated and compared with a historical benchmark value; when the energy consumption per unit volume of that layer exceeds a preset threshold, the current printing process parameters are automatically adjusted, and the adjusted parameters are applied starting from the next layer.

[0023] Specifically, the power consumption increment of this layer is:

[0024] In the formula: This indicates the power consumption value [kJ] displayed on the digital meter after the previous layer is completed;

[0025] The energy consumption per unit volume of this layer is:

[0026] ,

[0027] In the formula, This represents the cumulative power consumption displayed on the digital meter after the current layer is completed, in kJ. This represents the cumulative power consumption displayed on the digital meter after the previous layer is completed, in kJ. This represents the volume increment of the molten part in this layer, in mm³; the calculated... Compare with historical benchmarks; when When the preset threshold is exceeded, the current printing process parameters are automatically adjusted, and the adjusted parameters are applied starting from the next layer.

[0028] Preferably, the preset threshold is 1.5 times the historical benchmark value; the automatic adjustment of printing process parameters includes at least one of reducing laser power by 5%-15%, increasing scanning speed by 5%-15%, or increasing scanning spacing by 5%-10%.

[0029] Preferably, the total energy consumption of the printing process is allocated to each part according to the ratio of the sum of the finished volume, processing allowance volume and support structure volume of each part to the sum of the volumes of all parts in the batch.

[0030] The total energy consumption of the wire electrical discharge machining process is allocated to each part according to the ratio of the contact area between each part and the substrate to the sum of the contact areas of all parts in the batch.

[0031] The energy consumption of the printing process for the k-th part in the nth batch is allocated and calculated using the following formula:

[0032]

[0033] In the formula: The total energy consumption for printing the nth batch of parts is [kJ], which is obtained by reading a digital electricity meter. The volume of the k-th part in the nth batch [mm3] is calculated using 3D modeling software (such as Materialise Magics). The machining allowance volume [mm3] of the k-th part in the n-th batch is calculated using 3D modeling software (such as MaterialiseMagics). The volume of the supporting structure of the k-th part in the n-th batch is represented by [mm3], which is calculated using 3D modeling software (such as Materialise Magics); K is the total number of parts in this batch.

[0034] The energy consumption of the wire electrical discharge machining process for the k-th part in the nth batch is allocated according to the following formula:

[0035]

[0036] In the formula: The energy consumption [kJ] of the wire electrical discharge machining process for the k-th part in the nth batch; The total energy consumption [kJ] for wire EDM of the nth batch of parts is represented by the digital meter reading. This represents the contact area [mm] between the k-th part in the n-th batch and the substrate. 2 The calculation is performed using 3D modeling software such as Materialise Magics.

[0037] The total energy consumption of the automated guided vehicle (AGV) transportation process is allocated to each part according to the proportion of the product of the weight of each part and the transportation distance to the sum of the products of the weight of all parts and the transportation distance in the batch.

[0038] The system allocates and calculates the energy consumption of each part in the nth batch of transportation processes according to the following formula:

[0039]

[0040] In the formula: The energy consumption [kJ] of the AGV transportation process for the k-th part in the nth batch; The total energy consumption of the AGV transport of the nth batch of parts is represented by [kJ]. This represents the weight [g] of the k-th part in the n-th batch; This represents the transport distance [m] of the k-th part in the nth batch from the AGV starting point to the target CNC machine tool (which can be obtained from the workshop layout diagram).

[0041] The calculated The numerical value is input into the 3rd column and kth row of the nth dynamic energy consumption matrix.

[0042] Preferably, the energy consumption of the CNC finishing process is obtained by triggering the digital meter of the process to zero and read the data through an RFID tag attached to the part, thereby obtaining the finishing energy consumption of a single part. .

[0043] Preferably, the energy efficiency level is calibrated as A to E according to the ratio of energy consumption value to historical benchmark value, and the preset level is C.

[0044] Once all data in the nth dynamic energy consumption matrix is ​​filled, the values ​​in the four columns of the kth row are summed to obtain the total energy consumption of the kth part in the nth batch during the entire LPBF production process:

[0045]

[0046] Calculate the energy consumption per unit volume during the entire LPBF production process of this part:

[0047]

[0048] This Compared with historical benchmarks The energy efficiency rating of this part is determined by comparison according to the following rules:

[0049] Grade A: Grade B: Grade C: Grade D: Class E: .

[0050] Preferably, the improvement measures include at least one of optimizing the build direction, adjusting printing process parameters, regrouping, or optimizing cutting parameters; the method further includes storing the current batch of measured data into a historical database to update various benchmark values.

[0051] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0052] This invention achieves automated energy consumption data collection for each process from substrate to finished product under conditions where there are no physical parts before LPBF printing through the linkage triggering mechanism of RFID tags and digital meters in the physical sensing layer; through the dynamic energy consumption matrix in the data processing layer with parts as rows and processes as columns, the energy consumption data scattered in each process is collected by part, providing a data foundation for individual energy consumption traceability; through the control execution layer, the energy efficiency assessment results are automatically converted into adjustment instructions for the next batch of production plan, forming a closed-loop control of the entire process.

[0053] This invention overcomes the technical obstacle of not being able to directly measure the individual energy consumption of multiple parts printed on the same board through a "collection-allocation-feedback" method: printing energy consumption is allocated according to volume ratio, wire cutting energy consumption is allocated according to contact area ratio, and AGV energy consumption is allocated according to mass-distance weighted average, so that the total energy consumption of the batch can be reasonably collected to each part; the energy consumption value of each part is compared with historical benchmarks and the AE energy efficiency level is calibrated, realizing refined energy efficiency assessment at the three levels of part, process, and batch; the improvement measures corresponding to low energy efficiency are automatically applied to the next batch, realizing continuous energy-saving optimization between batches. Attached Figure Description

[0054] Figure 1 This is a flowchart of the method.

[0055] Figure 2 This is a system architecture diagram.

[0056] Marked in the image:

[0057] 100 - Physical sensing layer; 200 - Data processing layer; 300 - Control execution layer; 110 - First RFID tag; 120 - Second RFID tag; 130 - RFID reader / writer; 140 - Digital meter; 150 - Coulomb counter. Detailed Implementation

[0058] The present invention will now be described in further detail with reference to specific embodiments. However, this should not be construed as limiting the scope of the present invention to the following embodiments; all technologies implemented based on the content of the present invention fall within the scope of the present invention.

[0059] Example 1

[0060] like Figure 2 As shown, the system includes a physical sensing layer 100, a data processing layer 200, and a control execution layer 300.

[0061] The physical sensing layer 100 includes: a first radio frequency identification (RFID) tag 110 disposed on the substrate of the laser powder bed fusion (LPBF) printer; a second RFID tag 120 disposed on each part after cutting and separation; an RFID reader / writer 130 disposed at each production device; and a digital meter 140. The digital meter (including a coulomb counter for AGVs) is communicatively connected to the data processing layer via a wired network or wireless network (such as Wi-Fi or Bluetooth) to upload the collected power and energy consumption data in real time.

[0062] Specifically, the first RFID tag 110 is a high-temperature resistant metal-based RFID tag, capable of withstanding the high-temperature environment during the LPBF printing process. It is installed on each printer substrate to be used and stores a unique batch number corresponding to that substrate. The second RFID tag 120 is a general-purpose RFID tag, attached to each cut part after the wire EDM process, storing the part number of that part, used to trigger RFID read events at the corresponding equipment during each process flow. RFID readers 130 are deployed at each LPBF printer, wire EDM machine, automated guided vehicle (AGV), and CNC machine tool to read RFID tag information on the substrate or parts at the corresponding workstation. Digital meters 140 are connected to each LPBF printer, wire EDM machine, AGV, and CNC machine tool. The digital meters used for the AGV can be coulomb meters with wireless communication capabilities (e.g., H68B coulomb meters, which can communicate via Bluetooth).

[0063] The working principle of the physical sensing layer 100 is as follows: when a substrate or part carrying an RFID tag arrives at a certain process equipment, the RFID reader 130 at the equipment reads the RFID tag information and generates an RFID read event; the RFID read event triggers the digital meter 140 at the equipment to perform operations such as zeroing, starting recording or stopping recording, thereby realizing the collection and recording of energy consumption data of each process.

[0064] The data processing layer 200 is communicatively connected to each RFID reader 130 and digital meter 140 in the physical sensing layer 100. The data processing layer 200 receives energy consumption data for each process collected by the physical sensing layer 100 and constructs a dynamic energy consumption matrix for each batch, with parts as rows and processes as columns. The processes include at least four steps: LPBF printing, wire EDM, AGV transportation, and CNC finishing. The data processing layer 200 also allocates the total energy consumption of each process to each part according to a preset allocation model and fills the corresponding position in the dynamic energy consumption matrix, as well as calculating the energy consumption value of each part in each process and throughout the entire process based on the dynamic energy consumption matrix. Furthermore, the data processing layer 200 maintains a historical database, storing baseline energy consumption values ​​under various process parameters.

[0065] The control execution layer 300 is communicatively connected to the data processing layer 200. Based on a comparison between the energy consumption value calculated by the data processing layer 200 and historical benchmark values, it calibrates the energy efficiency level of each part, process, and batch. It then generates adjustment instructions for subsequent batch production plans based on the energy efficiency calculation results, which are executed through the control systems of each production equipment. The adjustment instructions include at least optimizing one of the following: part construction direction, printing process parameters (such as laser power, scanning speed, etc.), part batching plan, or cutting parameters.

[0066] In the above system, the physical sensing layer 100 and the data processing layer 200, and the data processing layer 200 and the control execution layer 300 can be connected via wired or wireless networks to realize real-time data transmission and the issuance of control commands.

[0067] Example 2

[0068] An energy-saving control method for a laser powder bed melting production process, implemented using the aforementioned system, specifically includes the following steps: Figure 1 As shown.

[0069] Step 1: RFID device deployment and multi-board batch binding

[0070] (a) Number each part in an additive manufacturing order and establish a "part number → order number" mapping. If one LPBF printer substrate cannot accommodate all parts in the order for printing at once, it needs to be distributed to multiple substrates in batches. Install a high-temperature resistant RFID tag on each substrate to be used and establish an "RFID tag → substrate number" mapping.

[0071] (b) The production equipment used in the entire LPBF production process includes several LPBF printers, wire EDM machines, AGV transport vehicles, and CNC machine tools. An RFID reader is deployed at each LPBF printer, wire EDM machine, AGV transport vehicle, and CNC machine tool, and an RFID reader → equipment number mapping is established.

[0072] (c) Connect a digital meter to each LPBF printer, wire EDM machine, AGV transport vehicle, and CNC machine tool, and establish a "digital meter → device number" mapping. The digital meter used for the AGV transport vehicle can be a coulomb meter with wireless communication function (e.g., an H68B coulomb meter, which can communicate via Bluetooth).

[0073] (d) Before production begins, the system generates a unique batch number for each substrate and writes it into the corresponding substrate's RFID tag. In addition, a batch number → order number mapping is established.

[0074] Step 2: Generate initial production plan (based on heuristic rules)

[0075] 2.1 Selection of Construction Direction

[0076] For each part to be printed, 27 representative feasible construction directions are generated by selecting discrete options at 30° intervals around the X, Y, and Z axes. 3D modeling software (such as Materialise Magics) is used to calculate the support structure volume for each construction direction, and the direction with the smallest support volume is selected as the preferred construction direction for the part.

[0077] 2.2 Component batching (based on substrate height)

[0078] (a) Arrange all parts in ascending order of their height (maximum dimension along the printing growth direction) from low to high.

[0079] (b) Assign the parts to the substrates sequentially, following the principle of assigning parts of similar height to the same substrate to minimize the height difference between parts on each substrate. This ensures that the printing completion time of each part is similar and avoids excessively extending the overall printing time due to a few tall parts. In addition, do not nest the parts; place only one layer of parts on each substrate.

[0080] (c) Specific allocation method: Set a height grouping threshold (e.g., 5 mm) to prioritize the allocation of parts whose heights fall within the same range to the same substrate; when the available area of ​​a substrate is insufficient to accommodate all parts in the group, the next substrate is activated. When a substrate still has extra area after accommodating all parts in the group, parts are selected from the next higher height range and added sequentially until no more parts can be accommodated, thereby improving the substrate area utilization rate.

[0081] (d) Record the part number and the build orientation of each part in each substrate (batch). Establish a "part number → batch number" mapping.

[0082] 2.3 Post-processing initial sequence

[0083] (a) The electrical discharge wire cutting sequence is performed from left to right according to the position of the part on the substrate.

[0084] (b) AGV transportation is carried out in whole batches. If multiple parts in a batch need to be transported to different CNC machine tools, the AGV will deliver them sequentially according to the planned route, and can deliver them to the nearest machine tool based on the distance of the parts to the target machine tool.

[0085] (c) The order of CNC machine tool processing of parts is sorted by part number from smallest to largest.

[0086] Step 3: Construct a dynamic energy consumption matrix

[0087] Build a dynamic energy consumption matrix for each batch of parts in the order. This is used to record the energy consumption of each part in each process step in real time. Matrix The matrix is ​​structured with part numbers as rows and process numbers as columns. The laser powder bed fusion (LPBF) production process involves four processes: LPBF printing, wire EDM, AGV transport, and CNC finishing. Therefore, the matrix has 4 columns. The first process is LPBF printing, the second is wire EDM, the third is AGV transport, and the fourth is CNC finishing. The number of rows in the matrix is ​​the number of parts (K) in the batch. The constructed matrix... as follows:

[0088]

[0089] The element in the k-th row and l-th column of the matrix represents the energy consumption value [kJ] of the k-th part in the l-th process of the n-th batch.

[0090] Step 4: Interlayer energy consumption monitoring in the LPBF printing process (core innovation)

[0091] When the RFID reader at the LPBF printer reads the RFID tag on the substrate, the RFID read event triggers the digital meter at the LPBF printer to zero. Then the printer prints all the parts on the substrate, and the digital meter collects power and energy consumption data in real time.

[0092] The first nine layers of printing are in the powder spreading and support structure establishment stage, during which energy consumption fluctuates significantly and does not represent a stable printing state; therefore, they are not monitored. Starting from the tenth layer, after each layer of powder melting is completed, the system executes the following sub-steps:

[0093] (a) After the current layer of powder has melted, the host computer reads the power consumption value displayed on the digital meter. Calculate the power consumption increment of this layer:

[0094]

[0095] In the formula: This indicates the power consumption value [kJ] displayed by the digital meter after the previous layer is completed.

[0096] (b) Obtain the volume increment of the molten part in that layer from the slice data. [mm3]. The slice data was obtained by slicing and processing the 3D model of the part using Materialise Magics software.

[0097] (c) Calculate the energy consumption per unit volume of this layer: The unit is J / mm³. Simultaneously, calculate the energy consumption per unit volume of the printed portion of the part: ,in This represents the cumulative volume [mm3] of all parts in the batch up to the current layer.

[0098] (d) This layer , Compare the energy consumption per unit volume for printing the same material using the same printer model in the historical database (using the P50 quantile). If... Exceeding 1.5 times the benchmark value or If the energy consumption exceeds 1.25 times the baseline value, the system determines that the energy consumption is abnormal. First, check if the LPBF printer is faulty. If it is faulty, repair it. If it is not faulty, the system will automatically perform at least one of the following adjustments: ① Reduce the laser power by 5%-15%; ② Increase the scanning speed by 5%-15%; ③ Appropriately increase the scanning spacing (increase by 5%-10%). The adjusted parameters should be within the process window to ensure the forming quality.

[0099] (e) The adjusted parameters are applied starting from the next layer of powder melting. The computer system records abnormal events (layer number, deviation, adjustment action) for subsequent self-learning.

[0100] (f) The above adjustments may extend the single-layer scanning time (e.g., after reducing power, the exposure time needs to be increased or the speed needs to be reduced to ensure the penetration depth), thereby increasing the total printing time. However, they can effectively reduce the energy consumption per unit volume, achieving a time-energy trade-off. Users can choose whether to enable this function based on their production priorities.

[0101] Repeat interlayer energy consumption monitoring until all layers are printed.

[0102] Step 5: Calculation of part-level energy consumption for LPBF printing process

[0103] After the nth batch of parts has been printed, the RFID reader at the LPBF printer reads the RFID tag on the substrate again. The system records the value of the digital meter at the LPBF printer at that moment, which represents the total energy consumption for printing that batch of parts. Then, calculate the energy consumption of the printing process for each part. The energy consumption of the printing process for the k-th part in the n-th batch is allocated and calculated using the following formula:

[0104]

[0105] In the formula: The total energy consumption for printing the nth batch of parts is [kJ], which is obtained by reading a digital electricity meter. The volume of the k-th part in the nth batch [mm3] is calculated using 3D modeling software (such as Materialise Magics). The machining allowance volume [mm3] of the k-th part in the n-th batch is calculated using 3D modeling software (such as MaterialiseMagics). The volume of the supporting structure of the k-th part in the n-th batch is represented by [mm3], which is calculated using 3D modeling software (such as Materialise Magics); K is the total number of parts in this batch.

[0106] The calculated The numerical value is input into the first column and kth row of the nth dynamic energy consumption matrix.

[0107] Repeat the above calculations until the first column of the matrix is ​​filled.

[0108] Step 6: Calculation of part-level energy consumption and RFID attachment for wire EDM process

[0109] After the LPBF printing process for the nth batch of parts is completed, the wire EDM process for that batch of parts begins. First, the RFID tag on the substrate is swiped once on the RFID reader / writer at the wire EDM machine, generating an RFID read event and triggering the digital meter at the wire EDM machine to zero. Then, the wire EDM machine cuts and separates all the parts on the substrate, while the digital meter collects power and energy consumption data in real time.

[0110] After all parts have been separated, the RFID reader at the wire EDM machine reads the RFID tag on the substrate again. The system records the value of the digital meter at the wire EDM machine at that moment, which is the total energy consumption for wire cutting of this batch of parts. .

[0111] An RFID tag is attached to each cut part, establishing a "part number → RFID tag" mapping. Each part's RFID tag is swiped once on the RFID reader at the wire EDM machine. The system allocates and calculates the energy consumption of the wire EDM process for each part corresponding to the RFID tag using the following formula:

[0112]

[0113] In the formula: The energy consumption [kJ] of the wire electrical discharge machining process for the k-th part in the nth batch; The total energy consumption [kJ] for wire EDM of the nth batch of parts is represented by the digital meter reading. The contact area [mm2] between the k-th part in the n-th batch and the substrate is represented by 3D modeling software (such as Materialise Magics).

[0114] The calculated The numerical value is input into the second column and the kth row of the nth dynamic energy consumption matrix.

[0115] Repeat the above calculations until the second column of the matrix is ​​filled.

[0116] Step 7: Calculation of component-level energy consumption in the AGV transportation process

[0117] After the wire EDM process of the nth batch of parts is completed, each part is weighed and its weight is recorded. Then, all parts in the batch are loaded onto an AGV (Automated Guided Vehicle). The RFID tag of the first loaded part is scanned on the RFID reader on the AGV. This RFID read event triggers the digital meter on the AGV to reset and begin recording the AGV's power consumption. The AGV then transports the parts to different types of CNC machine tools (e.g., lathes, milling machines, drilling machines, grinding machines, etc.) for finishing according to their processing requirements. When the last part in the nth batch is unloaded from the AGV, its RFID tag is scanned on the RFID reader on the AGV, generating an RFID read event. This triggers the host computer to read the digital meter value on the AGV at that moment, which represents the total energy consumption for transporting this batch of parts. .

[0118] The system allocates and calculates the energy consumption of each part in the nth batch of transportation processes according to the following formula:

[0119]

[0120] In the formula: The energy consumption [kJ] of the AGV transportation process for the k-th part in the nth batch; The total energy consumption of the AGV transport of the nth batch of parts is represented by [kJ]. This represents the weight [g] of the k-th part in the n-th batch; This represents the transport distance [m] of the k-th part in the nth batch from the AGV starting point to the target CNC machine tool (which can be obtained from the workshop layout diagram).

[0121] The calculated The numerical value is input into the 3rd column and kth row of the nth dynamic energy consumption matrix.

[0122] Repeat the above calculations until the third column of the matrix is ​​filled.

[0123] Step 8: Calculation of part-level energy consumption for CNC finishing process

[0124] After the AGV delivers the part to the designated machine tool, before the part is clamped onto the CNC machine tool, the RFID tag on the part is swiped once on the RFID reader / writer at the CNC machine tool, generating an RFID read event. This RFID read event triggers the digital meter at the CNC machine tool to reset to zero and begin recording the CNC machine tool's power consumption. After the part is finished on the CNC machine tool, the RFID tag on the part is swiped again on the RFID reader / writer at the CNC machine tool, generating another RFID read event. This RFID read event triggers the host computer to read the value from the digital meter at the CNC machine tool; this value represents the energy consumption of the finishing process for the k-th part in the nth batch. .

[0125] The measured The numerical value is input into the 4th column and kth row of the nth dynamic energy consumption matrix.

[0126] Repeat the above steps until the fourth column of the matrix is ​​filled.

[0127] Step 9: Energy consumption calculation and energy efficiency rating of the entire LPBF production process based on dynamic energy consumption matrix

[0128] Once all data in the nth dynamic energy consumption matrix is ​​filled, the values ​​in the four columns of the kth row are summed to obtain the total energy consumption of the kth part in the nth batch during the entire LPBF production process:

[0129]

[0130] Calculate the energy consumption per unit volume during the entire LPBF production process of this part:

[0131]

[0132] In the formula: This represents the unit volume energy consumption [J / mm3] of the entire LPBF production process for the k-th part in the nth batch. This value is compared with the benchmark unit volume energy consumption (P50 quantile) of the entire LPBF production process for the same material in the historical database. The retrieved benchmark value is... The energy efficiency rating of each part in the nth batch is determined according to the following rules:

[0133] Grade A: Grade B: Grade C: Grade D: Class E: .

[0134] Calculate the energy consumption per unit volume of the nth batch of parts in the LPBF printing process:

[0135]

[0136] This value was compared with the benchmark value (P50 quantile) for energy consumption per unit volume when printing the same material using the same LPBF printer in a historical database. The retrieved benchmark value was... The energy efficiency rating of the LPBF printing process for this batch of parts is determined according to the following rules:

[0137] Grade A: Grade B: Grade C: Grade D: Class E: .

[0138] The total energy consumption of the entire LPBF production process for the nth batch of parts is obtained by summing all K rows and 4 columns of the nth dynamic energy consumption matrix:

[0139]

[0140] Calculate the unit volume energy consumption of the entire LPBF production process for the nth batch of parts:

[0141]

[0142] Compare this value with By comparison, the energy efficiency rating of the entire LPBF production process for the nth batch of parts is determined according to the following rules:

[0143] Grade A: Grade B: Grade C: Grade D: Class E: .

[0144] For parts or processes achieving an energy efficiency rating of A, record the corresponding process design parameters and equipment numbers for reference in subsequent batches. For parts or processes with an energy efficiency rating lower than C (i.e., D or E), analyze the reasons and make improvements. Improvement measures include, but are not limited to: optimizing the build direction, adjusting printing process parameters (laser power, scanning speed, etc.), regrouping, and optimizing cutting parameters. The system can automatically apply the improvement suggestions to the initial production plan of the next batch, forming a closed-loop feedback optimization.

[0145] After the energy efficiency rating of the current batch is completed, the measured data will be stored in the historical database and various benchmark values ​​will be updated.

[0146] Based on the improvement suggestions, the control execution layer automatically adjusts the discrete interval of the candidate set of construction directions in step 2.1, or automatically adjusts the height grouping threshold in step 2.2, thereby generating a new, optimized initial production plan.

[0147] Step 10: End energy-saving control

[0148] Once all parts in the order have been manufactured and energy efficiency data has been recorded, the energy-saving control process for the entire LPBF production process of this order will end. Otherwise, return to step 3 to perform energy-saving control for the next batch of parts' LPBF production process.

[0149] Specific Case Analysis

[0150] The following is combined Figure 1 The system architecture shown, taking the mass production of automotive turbocharger impellers (AlSi10Mg material) as an example, provides a detailed description of the specific implementation of the energy-saving control method for the laser powder bed melting production process provided by this invention. The data in this embodiment is based on measured values ​​from the CAD model of the WF model impeller and data collected on-site, aiming to reflect a real production scenario. However, the method of this invention is also applicable to other metallic materials (such as Ti6Al4V, Inconel 718, 316L, etc.) and other mass-produced LPBF parts with similar morphologies.

[0151] Orders and Equipment Configuration

[0152] Order DD20260519007 requirements:

[0153] 107 turbocharger impellers were produced, made of AlSi10Mg material with a density of 2.677 g / cm³. 3 = 0.002677 g / mm 3 Each impeller has a different finished volume V (measured directly from the CAD model), machining allowance volume Z (including shrinkage compensation and subsequent cutting), and support structure volume S (determined by the construction direction).

[0154] According to statistics: ① Finished product volume V: range 34418 mm 3 - 35153 mm 3 Average value 34792 mm 3 ② Machining allowance Z: range 1702 mm 3 - 1978 mm 3 Average value 1846 mm 3 ③ Support volume S: range 6781 mm 3 - 7342 mm 3 A few reached 8967 mm due to complex curved surfaces. 3 - 9430 mm 3 .

[0155] The equipment configuration is as follows: LPBF printer: Model SLM 280 HL (Dual laser 400W), substrate size 280 mm × 280 mm, measured average power during printing stage 3.08 kW. ② Wire EDM machine: Model DK7740, average power during wire cutting process 1.52 kW. ③ AGV trolley: Model CH-MG-850, magnetic navigation, average travel power 0.51 kW, equipped with H68B coulomb meter. ④ CNC machine tool: Model VMC1060, three-axis vertical milling machine, average power 2.95 kW. ⑤ RFID system: each substrate is equipped with a high-temperature resistant metal-based RFID tag; RFID readers are deployed at the entrance of each process equipment; an automatic labeling machine is equipped at the wire cutting machine exit.

[0156] Step 1: RFID device deployment and multi-board batch binding

[0157] The system numbered each impeller from T001 to T107, establishing a mapping between "part number → order number DD20260519007". After layout optimization, a single base plate (280 × 280 mm²) can accommodate a maximum of 36 impellers, therefore they are divided into three batches: base plate A (batch number B01), base plate B (batch number B02), and base plate C (batch number B03). An RFID tag with the batch number is installed on each of base plates A, B, and C. An RFID reader / writer and a digital meter are deployed at the LPBF printer, wire cutting machine, AGV, and CNC machine tool.

[0158] Step 2: Generate the initial production plan

[0159] 2.1 Selection of Construction Direction

[0160] For each impeller to be printed in the order, 27 representative feasible construction directions were generated by selecting impellers at 30° intervals around the X, Y, and Z axes. The support volume for each direction was calculated using Materialise Magics software. The results show that: ① The support volume of 98 impellers is smallest in the "shaft hole end face downward" direction, with a measured range of 6781 mm3 - 7342 mm3 and an average of 7079 mm3. ② The support volumes of 6 impellers (T020, T035, T053, T068, T082, and T104) are smallest after rotating them 60° around the X-axis due to the special concavity on the back of the blades, with measured volumes of 9125 mm3, 8967 mm3, 9274 mm3, 9430 mm3, 9311 mm3, and 8983 mm3, respectively. ③ The three impellers (T014, T060, T091) have the smallest supporting volumes after rotating 120° around the Y-axis (6891 mm3, 7043 mm3, 6955 mm3, respectively). The system records the preferred construction direction for each impeller.

[0161] 2.2 Parts batching

[0162] All 107 impellers were arranged in ascending order of height (height range 34.5 - 42.6 mm), with a height grouping threshold of 4 mm. 36 impellers with heights of 34.5 - 38.5 mm were assigned to substrate A, 36 impellers with heights of 38.6 - 42.6 mm to substrate B, and 36 impellers with heights of 42.7 - 46.7 mm to substrate C. After arranging the parts on substrate A using the lowest horizontal line method, the occupied area was 276.5 mm × 275.2 mm, leaving insufficient space to accommodate another complete part. The impeller numbers and corresponding construction orientations of the parts on substrate A were recorded.

[0163] 2.3 Post-processing initial sequence

[0164] Wire cutting sequence: Perform the cutting from left to right according to the position of the parts on substrate A.

[0165] AGV transportation: The entire batch of 36 impellers was loaded and transported at once.

[0166] CNC machining sequence: sorted by part number from smallest to largest.

[0167] Step 3: Construct a dynamic energy consumption matrix

[0168] Construct a 36-row × 4-column dynamic energy consumption matrix for substrate A (batch number B01). The rows correspond to parts T001-T036, and the columns correspond to the processes: Column 1 LPBF printing, Column 2 wire EDM, Column 3 AGV transportation, and Column 4 CNC finishing. All elements are initialized to 0.

[0169] Step 4: Monitoring interlayer energy consumption in the LPBF printing process

[0170] Substrate A (batch number B01) is loaded into the printer. The RFID reader reads the RFID tag on substrate A, triggering the digital meter to zero. The printer begins printing, and the digital meter starts recording power consumption. Process parameters: laser power 370 W, scanning speed 1200 mm / s, layer thickness 30 μm, scanning spacing 0.12 mm. Total layers: 1325 (maximum part height 38.5 mm). According to actual measurements, the printer's average power during the printing stage is 3.08 kW, the total printing time is 62280 s, and the theoretical total energy consumption is: 3.08 × 62280 = 191822 kJ.

[0171] No monitoring is performed on the first 9 layers (the powdering and support establishment stage).

[0172] Starting from the 10th layer, after each layer of powder melting is completed, the following steps are performed: ① Read the current cumulative power consumption of the digital meter. ① Calculate the increment of this layer; ② Obtain the volume increment of the molten part in this layer from the Materialise Magics slice data. ③ Calculate the energy consumption per unit volume of this floor. Energy consumption per unit volume of the printed part ④ The baseline value (P50 quantile) for the energy consumption per unit volume of AlSi10Mg material printed by the SLM 280HL printer in the historical database is 132.3 J / mm3.

[0173] When printing reaches layer 362, the calculation is obtained. The energy consumption exceeded the benchmark value by 1.5 times, indicating abnormal energy consumption. The printer was found to be functioning correctly, but the large support volume of this layer caused the problem. The power increased significantly, and the system automatically reduced the laser power from 370W to 336W (a reduction of 9.2%).

[0174] Adjusted average of floors 363-500 The concentration dropped to 162.8 J / mm³, 1.25 times lower than the baseline, and no further adjustment was triggered. The average concentration from layer 501 to layer 1325... It is 118.7 J / mm3, which is 10.3% lower than the benchmark value.

[0175] Printing finished, the digital meter displays the total energy consumption. (51.53 kWh, about 3.3% lower than the theoretical value).

[0176] Step 5: Calculation of part-level energy consumption in the printing process

[0177] The 36 impellers in substrate A were extracted using Materialise Magics software. , and Summation: Therefore, the average finished volume, machining allowance volume, and support structure volume of each impeller sum to approximately 43,834 mm³. The printing energy consumption for each part is calculated based on volume allocation, as shown in Table 1.

[0178]

[0179] Table 1 Energy Consumption Allocation Data for Each Impeller Printing Process on Substrate A

[0180] The following are some representative component data:

[0181] Each part Fill in the dynamic energy consumption matrix The first column corresponds to the row.

[0182] Step 6: Calculation of part-level energy consumption and RFID attachment for wire EDM process

[0183] After substrate A cools, it is moved to the wire EDM machine. First, the RFID tag on substrate A is swiped once on the RFID reader at the wire EDM machine, resetting the corresponding digital meter to zero. Then, the 36 impellers are cut and separated one by one. Once all impellers have separated from substrate A, the RFID reader reads the RFID tag on the substrate again, and the system automatically records the digital meter reading at that moment. Each impeller is automatically labeled after cutting (RFID tag with part number written on it) and swiped once on a reader. The total contact area between substrate A and the 36 impellers is calculated to be 8530 mm² using Materialise Magics software. The energy consumption of the wire EDM process for each impeller is distributed according to its contact area:

[0184]

[0185] For example, given that the contact area of ​​impeller T001 is 243 mm², the energy consumption of the wire EDM process for this impeller is: .

[0186] Calculate the values ​​of each impeller sequentially. And fill in the dynamic energy consumption matrix. The second column corresponds to the row.

[0187] Step 7: Calculation of component-level energy consumption in the AGV transportation process

[0188] All 36 impellers were loaded onto the AGV (Automated Guided Vehicle). The RFID tag of the first loaded part was swiped on the RFID reader on the AGV, triggering the coulomb counter to zero and begin recording. The AGV transported the impellers to the CNC machining area along a preset path, with the coulomb counter accumulating power consumption in real time during transport. When the last part was unloaded from the AGV, its RFID tag was swiped on the RFID reader on the AGV, triggering the host computer to read the coulomb count value, thus obtaining the measured total energy consumption for the AGV transport of this batch of parts. Calculated based on the mass of each impeller and the transport distance: The energy consumption of each AGV component is weighted by mass and distance:

[0189]

[0190] For example, given that the impeller T001 has a mass of 116.3 g and a transport distance of 121 meters, the energy consumption of the AGV transport process for this impeller is: .

[0191] Calculate the values ​​of each impeller sequentially. And fill in the dynamic energy consumption matrix. The corresponding row in column 3.

[0192] Step 8: Calculation of part-level energy consumption for CNC finishing process

[0193] After the AGV delivers the impeller part to the designated CNC machine tool, before clamping the part, it swipes the RFID tag on the part onto the RFID reader on the CNC machine tool, generating an RFID read event. This event triggers the digital meter on the CNC machine tool to zero and begins recording the power consumption during the part's machining process. After the part is finished on the CNC machine tool, the RFID tag is swiped again onto the RFID reader, generating another RFID read event. This triggers the host computer to read the digital meter value, which represents the energy consumption of the finishing process for the k-th part in the first batch. The measured energy consumption results for CNC finishing of some parts in this batch are as follows: , , , , .

[0194] Calculate the values ​​of each impeller sequentially. And fill in the dynamic energy consumption matrix. The corresponding row in column 4.

[0195] Step 9: Energy consumption calculation and energy efficiency rating of the entire LPBF production process

[0196] Dynamic energy consumption matrix After all data is filled in, calculate the total energy consumption of the entire LPBF production process for each part within the batch. and energy consumption per unit volume Historical database of energy consumption per unit volume for similar parts 3. Data for some representative parts are shown in Table 2 below:

[0197] Table 2 shows the energy consumption and energy efficiency rating of a representative impeller on substrate A throughout the entire process.

[0198] The energy efficiency rating of parts T020 and T035 is D, and the reasons for this need to be analyzed and improvements made. The main reason is that the volume of the support structure for these two impellers is too large. It is recommended that more build directions be generated in the next batch for comparison and selection to reduce the support volume. In addition, the volume of the support structure itself should also be optimized.

[0199] After the energy efficiency rating of batch B01 was completed, the measured data of 36 impellers were stored in the historical database and various benchmark values ​​were updated.

[0200] Step 10: End energy-saving control

[0201] Order "DD20260519007" still has impellers T036 - T105 that have not been completed. Return to step 3 to perform energy-saving control throughout the LPBF production process of batch B02 impellers. It is recommended to optimize the construction direction of this batch of impellers first.

[0202] Overall effect

[0203] This embodiment fully demonstrates the entire process of the method of the present invention, from RFID deployment, initial scheme generation, inter-layer monitoring, energy consumption allocation, energy efficiency rating calibration to closed-loop feedback. In the first batch, four impeller parts had an energy efficiency rating of D due to poor selection of construction direction and poor support structure design. Through automatic optimization by the system, the energy consumption of subsequent batches was significantly reduced, verifying the effectiveness of the method.

[0204] This embodiment uses AlSi10Mg as an example, but the method of the present invention is also applicable to other metal materials (such as Ti6Al4V, Inconel 718, 316L, etc.) and other mass-produced LPBF parts with similar morphology.

[0205] This invention enables traceability of individual energy consumption throughout the entire process from powder raw materials to the final product, filling the technological gap in additive manufacturing where individual energy consumption cannot be directly measured. Through real-time inter-layer monitoring and adaptive adjustment, energy consumption per unit volume can be effectively reduced by sacrificing a small amount of printing time. Through closed-loop feedback optimization between batches, the system possesses self-learning and continuous improvement capabilities, resulting in a significant reduction in overall energy consumption.

[0206] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An energy-saving control system for a laser powder bed melting production process, characterized in that, include: The physical sensing layer includes a first RFID tag disposed on the substrate of the laser powder bed fusion printer, RFID readers and writers disposed at each production device, and a digital electricity meter. The digital electricity meter is signal-connected to the RFID readers and writers and is triggered when the RFID readers and writers generate a read event to collect energy consumption data of the corresponding process. Each production device includes at least a laser powder bed fusion printer, an EDM wire cutting machine, an automated guided vehicle, and a CNC machine tool. The data processing layer, which is communicatively connected to the physical sensing layer, is used to construct a dynamic energy consumption matrix with parts as rows and processes as columns. The processes include at least laser powder bed fusion printing, electrical discharge wire cutting, automated guided vehicle transportation, and CNC precision machining. The control execution layer, which is communicatively connected to the data processing layer, is used to generate and execute adjustment instructions for subsequent batch production plans based on the energy efficiency calculation results of each part or process in the dynamic energy consumption matrix.

2. The energy-saving control system for the laser powder bed melting production process according to claim 1, characterized in that, The first RFID tag is a high-temperature resistant metal-based tag that stores the batch number corresponding to the substrate. The system also includes a second RFID tag, which is attached to each part after cutting and separation. The second RFID tag stores the part number of the part and is used to trigger an RFID read event at the corresponding equipment during each process flow, so as to trigger the digital meter to collect and record energy consumption data.

3. The energy-saving control system for the laser powder bed melting production process according to claim 1, characterized in that, The adjustment instructions generated by the control execution layer include at least one of the following: optimization of part construction direction, printing process parameters, part batching scheme, or cutting parameters.

4. An energy-saving control method for a laser powder bed melting production process, characterized in that, The system described in any one of claims 1-3 includes the following steps: The radio frequency identification (RFID) read event in the physical sensing layer triggers the digital meter of the corresponding process to collect and record the total energy consumption of that process. Using the data processing layer, the total energy consumption of each process is allocated to each part and filled into the corresponding position of the dynamic energy consumption matrix. Based on the dynamic energy consumption matrix, the energy consumption value of each part in each process and the entire process is calculated. Using the control execution layer, the calculated energy consumption value is compared with the historical benchmark value. Based on the comparison result, the energy efficiency level is calibrated, and the improvement measures corresponding to the parts or processes with energy efficiency levels lower than the preset level are automatically applied to the production plan of the next batch.

5. The energy-saving control method according to claim 4, characterized in that, The printing process also includes an interlayer energy consumption monitoring step: after each layer of powder melting is completed, the energy consumption per unit volume of that layer is calculated and compared with the historical benchmark value; when the energy consumption per unit volume of that layer exceeds the preset threshold, the current printing process parameters are automatically adjusted, and the adjusted parameters are applied from the next layer onwards.

6. The energy-saving control method according to claim 5, characterized in that, The preset threshold is 1.5 times the historical benchmark value; the automatic adjustment of printing process parameters includes at least one of reducing laser power by 5%-15%, increasing scanning speed by 5%-15%, or increasing scanning spacing by 5%-10%.

7. The energy-saving control method according to claim 4, characterized in that, The total energy consumption of the printing process is allocated to each part according to the ratio of the sum of the finished volume, machining allowance volume and support structure volume of each part to the sum of the above three volumes of all parts in the batch. The total energy consumption of the wire electrical discharge machining process is allocated to each part according to the ratio of the contact area between each part and the substrate to the sum of the contact areas of all parts in the batch. The total energy consumption of the automated guided vehicle (AGV) transportation process is allocated to each part according to the proportion of the product of the weight of each part and the transportation distance to the sum of the products of the weight of all parts and the transportation distance in the batch.

8. The energy-saving control method according to claim 4, characterized in that, The energy consumption of the CNC precision machining process is obtained by identifying the radio frequency identification tag attached to the part, which triggers the zeroing and reading of the digital meter for that process, thereby obtaining the process energy consumption of the individual part.

9. The energy-saving control method according to claim 4, characterized in that, The energy efficiency level is calibrated as A to E based on the ratio of energy consumption value to historical benchmark value, and the preset level is C.

10. The energy-saving control method according to claim 4, characterized in that, The improvement measures include at least one of optimizing the build direction, adjusting printing process parameters, regrouping, or optimizing cutting parameters; the method also includes storing the current batch of measured data into a historical database to update various benchmark values.

Images