Multi-objective optimization method and integrated system based on numerical control gear grinding process database

Abstract

The invention relates to the technical field of numerical control machining parameter processing, in particular to a multi-objective optimization method based on a numerical control gear grinding process database and an integrated system. According to the optimization method, the rated parameters of the machine tool and the real-time environment state are obtained, the safety technological parameter range considering the power, the thermal load and the movement stroke is dynamically calculated, the mathematical model of the gear grinding efficiency, precision and surface quality is constructed in the range, and finally the optimal technological parameters capable of optimizing multiple project targets at the same time are obtained. The technical problem that the machining quality is unstable due to the fact that a current worm grinding wheel gear grinding process optimization method neglects the dynamic working condition of a machine tool and the environmental influence and lacks multi-target collaborative optimization on efficiency, precision and surface quality is effectively solved.

Description

Technical Field

[0001] This invention relates to the field of CNC machining parameter processing technology, specifically a multi-objective optimization method and integrated system based on a CNC gear grinding process database. Background Technology

[0002] In the field of CNC gear grinding, the setting of process parameters is a key factor determining machining efficiency, gear accuracy, and surface quality. Traditional worm wheel grinding process optimization methods usually rely on engineers' experience or fixed process manuals. These parameters are often set under ideal conditions based on the rated performance of the machine tool, lacking dynamic adaptability to actual machining conditions.

[0003] With the increasing demand for high-precision gears, especially in high-tech fields such as aerospace and new energy vehicles, more stringent requirements are being placed on the quality, efficiency, and stability of gear grinding. Existing process parameter optimization methods mainly have the following limitations: First, the actual state of the machine tool and the influence of the environment are ignored. Existing technology usually assumes that the machine tool always operates under rated power and ideal environmental conditions. However, in actual production, dynamic factors such as the machine tool's thermal load (e.g., spindle temperature rise, ambient temperature fluctuations), mechanical wear, and power grid voltage fluctuations can significantly change the machine tool's real-time machining capabilities. A process set under rated parameters may lead to a decrease in machining accuracy or even cause equipment overload or workpiece burns in a dynamically changing environment.

[0004] Second, the optimization objectives are singular, making it difficult to achieve multi-objective synergy. Many existing optimization methods focus on a single objective, such as pursuing the highest processing efficiency or ensuring a specific accuracy index. However, the optimization of gear grinding is essentially a complex problem that requires balancing multiple conflicting objectives such as efficiency, accuracy, and surface quality. For example, increasing grinding speed can improve efficiency, but may exacerbate thermal load, thereby affecting tooth profile accuracy and deteriorating surface quality. Existing methods lack a comprehensive mathematical model that can simultaneously quantify and synergistically optimize these three key objectives.

[0005] Third, the safety boundaries for process parameters are rigidly defined. Existing methods determine the safe range of process parameters primarily based on the machine tool's rated power and theoretical stroke, failing to consider real-time environmental conditions to calculate the actual capability boundaries under current operating conditions. Such static boundary settings are either too conservative, failing to fully utilize the equipment's potential, or too aggressive under harsh operating conditions, leading to processing risks.

[0006] Therefore, there is an urgent need in this field for a gear grinding process optimization method that can overcome the above-mentioned shortcomings. This method should be able to consider the machine tool status and external environment in real time, accurately define the safe operating range of process parameters, and, based on this, collaboratively optimize gear grinding efficiency, accuracy, and surface quality through a scientific mathematical model, thereby achieving intelligent, adaptive, and high-quality gear grinding. Summary of the Invention

[0007] To address the technical problem of unstable machining quality caused by neglecting the dynamic operating conditions of machine tools and the influence of the environment, and the lack of multi-objective collaborative optimization of efficiency, accuracy, and surface quality, this invention provides a multi-objective optimization method based on a CNC gear grinding process database. Based on this optimization method, this invention also provides an integrated system for implementing the optimization method.

[0008] To achieve the above objectives, the present invention provides the following technical solution: A multi-objective optimization method based on a CNC gear grinding process database is characterized by the following steps: S1. Obtain the rated machining parameters of the machine tool and the environmental conditions during gear grinding; S2. Based on environmental conditions, determine the capability boundaries of the computer-controlled gear grinding machine in terms of power, thermal load, and motion stroke, and refer to the rated machining parameters to determine the safe operating range of the gear grinding process parameters; S3. Within the safe operating range, obtain the three categories of measurement standards corresponding to each group of grinding process parameters: grinding efficiency, grinding accuracy, and surface quality, and construct a mathematical model to predict grinding efficiency, grinding accuracy, and surface quality. S4. Based on the mathematical model, obtain the target gear grinding process parameters corresponding to the target measurement standard, so as to guide the gear grinding process of the target machine tool.

[0009] As a further aspect of the present invention: the environmental conditions include the temperature range, humidity range, and coolant flow rate and temperature of the environment in which the grinding machine is located during gear grinding.

[0010] As a further aspect of the present invention: the safe operating range refers to the intersection of various gear grinding process parameter sets that respectively satisfy the capability boundaries in terms of power, thermal load, and motion stroke.

[0011] As a further aspect of the present invention, the mathematical model refers to a second-order response surface model consisting of grinding process parameters as independent variables and grinding efficiency, grinding accuracy, and surface quality as dependent variables.

[0012] As a further aspect of the present invention: in S4, a hybrid optimization algorithm combining genetic algorithm and particle swarm optimization algorithm is used to generate multiple candidate grinding process parameter schemes, and a candidate grinding process parameter scheme that meets the target evaluation criteria is selected.

[0013] As a further aspect of the present invention: after obtaining the target gear grinding process parameters, the parameters are sent to a CNC machine tool, including the following steps: Step 1: Establish a correspondence table between gear grinding process parameters and internal variables of the CNC system; Step 2: Based on the correspondence table, write the target gear grinding process parameters obtained by the above optimization method into the CNC system.

[0014] As a further aspect of the present invention: in step one, the correspondence table includes: Process parameter identifier, used to uniquely identify a gear grinding process parameter; The CNC variable path is used to specify the storage address or variable name of the gear grinding process parameters in the target CNC system; The unit scaling factor is used to define the conversion relationship between the units of the gear grinding process parameters and the units inside the CNC system. Read / write permission identifier, used to declare the read / write attributes of gear grinding process parameters in the CNC system; Parameter verification rules are used to verify the legality or logical consistency of parameter values ​​before writing the target gear grinding process parameters into the CNC system.

[0015] As a further aspect of the present invention: In step two, before batch writing process parameters to the CNC system, a desired consistency fingerprint is pre-calculated based on the target gear grinding process parameters; after batch writing the target gear grinding process parameters to the CNC system, all written parameter values ​​are immediately read back to form a readback parameter set; based on the readback parameter set, a readback consistency fingerprint is calculated; the readback consistency fingerprint is compared with the desired consistency fingerprint; if the two are consistent, the target gear grinding process parameters are confirmed to have been successfully issued and the parameter settings are submitted; if the two are inconsistent, a rollback operation is triggered to restore the parameters of the CNC system to the state before writing.

[0016] As a further aspect of the present invention: both the expected consistency fingerprint and the readback consistency fingerprint are verification codes calculated by applying a cryptographic hash function to the complete parameter dataset; the complete parameter dataset corresponding to the expected consistency fingerprint includes the complete and final command data packet to be written into the CNC system; the complete parameter dataset corresponding to the readback consistency fingerprint includes the complete and actual parameter status data packet read back from the CNC system.

[0017] This invention also provides a multi-objective optimization integrated system based on a CNC gear grinding process database. This system is used to implement the above-mentioned optimization method, including: The parameter mapping module is used to store and manage the correspondence table between the process parameters and the internal variables of the CNC system; The transaction control module is used to execute the transactional distribution steps, including compliance checks, batch writing, readback comparison, and result confirmation.

[0018] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention constructs a closed-loop intelligent decision-making system from "machine tool status perception" to "process parameter optimization". First, by synchronously acquiring the rated parameters of the machine tool and the real-time environmental status, it breaks through the limitation of traditional process optimization that only considers the static equipment capabilities, enabling the system to perceive dynamic working conditions. Second, it comprehensively calculates the real-time safety boundary from three dimensions: power, thermal load, and motion stroke. This avoids the waste of equipment efficiency caused by conservative parameters and prevents precision degradation and equipment damage caused by overload processing. Finally, within the safety boundary, a multi-objective prediction model for efficiency, accuracy, and surface quality is established. Through mathematical models, the process parameter selection that originally relied on experience is transformed into a quantifiable optimization problem. Ultimately, it outputs the optimal combination of process parameters that can simultaneously balance processing efficiency and product quality under the current working conditions, realizing a qualitative change from "fixed parameters" to "adaptive parameters". This effectively solves the technical problem of unstable processing quality caused by ignoring the dynamic working conditions and environmental influences of the machine tool and lacking multi-objective collaborative optimization of efficiency, accuracy, and surface quality.

[0019] 2. The standardized mapping mechanism between the worm gear grinding process parameters and the internal variables of the CNC system established in this invention achieves seamless distribution of optimization parameters through a predefined correspondence table, avoiding possible misoperation or data conversion errors caused by manual input, and ensuring that the optimization results can be accurately and completely transmitted to the execution end, thus ensuring the closed-loop reliability from algorithm optimization to machine tool execution.

[0020] 3. By refining the structure of the correspondence table, sophisticated management and security control of parameter distribution are effectively achieved. Specifically, the "numerical control variable path" ensures the accuracy of parameter writing location, the "unit scaling factor" eliminates the risk of unit mismatch, the "read / write permission identifier" prevents accidental modification of critical system parameters, and the "parameter verification rules" intercept illegal or logically conflicting values ​​before writing. This multi-layered protection mechanism at the underlying data level greatly improves the accuracy of parameter distribution and system stability.

[0021] 4. By comparing the "expected consistency fingerprint" and the "readback consistency fingerprint", this invention can not only detect whether anomalies occur during parameter transmission, but also automatically restore the system to a safe state through the transaction rollback mechanism, forming an industrial-grade secure distribution process with self-diagnosis and fault tolerance capabilities.

[0022] 5. This invention transforms the methodology into a concrete system implementation. The "parameter mapping module" ensures centralized management and traceability of parameter mapping, while the "transaction control module" enables atomic parameter distribution. This systematic design allows optimization parameters to be deployed in the industrial field with high reliability and repeatability, significantly reducing the processing quality risks caused by parameter distribution failures. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the overall system architecture of the present invention.

[0024] Figure 2 This is the data model and relationship diagram of the present invention.

[0025] Figure 3 This is a flowchart illustrating the database setup and instantiation process of this invention.

[0026] Figure 4 This is a diagram of the RSM-based multi-objective optimization process and Pareto file structure of the present invention.

[0027] Figure 5 This is a timing diagram for the domestic CNC mapping, transactional distribution, and readback of the present invention. Detailed Implementation

[0028] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] I. Multi-objective optimization method for worm gear grinding process parameters Semantic data modeling and template definition are employed to construct a complete set of template parameters covering gear features, machine tool capabilities, grinding wheels and dressing, machining strategies, cooling, and measurement elements. Changes are recorded incrementally using the process template as a baseline. Constraint consistency is achieved through context snapshots to generate initial parameter values ​​and a safety domain. Context snapshots consist of keys such as machine tool ID, grinding wheel ID, dressing count, and environmental interval. In subsequent optimization, Design of Effect (DOE) sampling is conducted within the safety domain. Experiments are designed and second-order response surface models for efficiency, accuracy, and texture are constructed. Based on this, a hybrid form of particle swarm optimization and genetic algorithm (GA-PSO) is used to solve the Pareto non-dominated solution set. Penalty and elimination strategies for power, grinding wheel wear, and stroke constraints are superimposed to obtain executable solutions, forming Pareto profiles with context keys and evidence fingerprints. Online solution selection and CNC output are then achieved through knee point recognition or preference weight rules.

[0030] This embodiment uses a 20CrNiMo carburized and quenched gear with a module m=3mm (range 0.8-8), number of teeth z=50 (range 12-200), and tooth width b=30mm (range 5-80mm) as the machining object. The target machine tool is a domestic CNC worm gear grinding machine (spindle rated power 18kW). Figures 1-4 Detailed implementation steps.

[0031] (a) Obtaining machine tool rated parameters and environmental conditions 1. Obtain the rated machining parameters of the CNC machine tool through the CNC system parameter interface (GetMachineParam function) and the machine tool manual, specifically including: Power parameters: Spindle rated power P_max=18kW, overload protection threshold P_ol=25kW, transmission efficiency η=0.88.

[0032] Thermal load parameters: maximum allowable spindle temperature T_spindle_max=85℃, maximum feed axis temperature T_axis_max=65℃, cooling system heat exchange power Q=15kW.

[0033] Motion stroke: X-axis ±300mm, Y-axis 0-400mm, Z-axis 0-500mm, B-axis (workpiece axis) positioning accuracy ±5″, C-axis (grinding wheel axis) maximum speed 4000r / min.

[0034] Grinding wheel compatibility parameters: diameter range φ is 200-400mm (in this embodiment, a 300mm CBN grinding wheel with a grit size of 120# and a binder resin is selected), dressing cycle is 50 pieces / time, and the maximum linear velocity v_s_max=45m / s.

[0035] The basic process parameters for gear grinding are: feed rate v_f = 0.3-2.0 mm / s, depth of cut a_e = 0.005-0.04 mm, and dwell time dwell = 0-2.0 s.

[0036] 2. Environmental status parameters are collected in real time through the workshop Internet of Things (IoT) system, specifically including: Ambient temperature and humidity: Temperature 22±3℃, humidity 50±10%RH.

[0037] Coolant parameters: Model: emulsion EC-5, concentration: 6%, flow rate: 80L / min, inlet temperature: 23±2℃.

[0038] Auxiliary parameters: workshop power grid voltage 380±10V, vibration level ≤0.05mm / s².

[0039] (ii) Obtaining the scope of safe use 1. According to Figure 2The steps involve classifying and entering the above parameters into the corresponding tables in the database. The core relationships are as follows: The machine_tool table (machine tool capabilities) stores parameters such as P_max, stroke, and speed. The primary key is machine_id=GMC2000.

[0040] The grinding_wheel table (grinding wheel profile) stores information such as grinding wheel diameter, grit size, and dressing cycle. The primary key is wheel_id=CBN300-120.

[0041] The context_snapshot table is associated with machine_id=GMC2000 and wheel_id=CBN300-120, storing environmental temperature and humidity, and coolant parameters. The primary key is context_id=CTX2024051001.

[0042] The gear_feature table (gear features) stores parameters such as m=3, z=50, b=30, and the primary key is gear_id=Gear20240510.

[0043] 2. Database table structure setup The database is built using SQLite, and the core table structure and fields are as follows: The gear_feature table contains: gear_id (primary key), m (0.8-8), z (12-200), α (20°±5°), β (0-35°), b (5-80mm), material, heat_treatment, and profile_mod.

[0044] machine_tool table: machine_id (primary key), P_max (kW), P_ol (kW), X_stroke (mm), Y_stroke (mm), Z_stroke (mm), C_max_rpm (r / min), T_spindle_max (℃).

[0045] The grinding_wheel table contains: wheel_id (primary key), D (mm), B (mm), abrasive (material), grain_size (grain size), bond (bond), and dressing_cycle (dressing cycle). The context_snapshot table contains: context_id (primary key), machine_id, wheel_id, coolant_flow (L / min), temp_range (°C), humidity_range (%RH), and checksum (verification fingerprint).

[0046] The process_template table contains: template_id (primary key), v_s_range (m / s), v_f_range (mm / s), a_e_range (mm), dwell_range (s), and dressing_param (trimming parameter). The process_instance_delta table contains: instance_id (primary key), template_id, Δparam (parameter increment), change_reason (reason for change), operator (operator), and timestamp (time stamp).

[0047] 3. Process template definition and instance incremental recording Template definition: Create a standard template (template_id=TEMP001) in the process_template table, with the following parameter ranges: v_s=25-45m / s, v_f=0.3-2.0mm / s, a_e=0.005-0.04mm, and dwell=0-2.0s.

[0048] Incremental Example: Since the gear in this embodiment is made of carburized and quenched material (hardness HRC60-62), the upper limit of a_e needs to be adjusted to 0.03mm. In the process_instance_delta table, record Δa_e=-0.005mm, instance_id=INS001, and associated template_id=TEMP001.

[0049] Constraint Consistency Solving and Security Domain Generation Figure 3 The process invokes the constraint consistency solver to complete five types of constraint checks and finally determines the safe usage range: Power boundary: Based on the cutting power formula P(t)=k_c•a_e•v_f•B•η -1 (k_c=2100N / mm) 2 (B=30mm), the calculated value is v_f≤1.4mm / s; Heat load boundary: Based on the heat balance equation ΔT=(P(t)-Q)•t / (m•c) (m=500kg, c=450J / (kg•℃)), continuous processing ≤40min; Motion travel boundaries: X-axis ±40mm, Z-axis ≤60mm; Grinding wheel wear boundary: single wear ≤ 0.01mm, cumulative a_e ≤ 0.15mm; Environmental compatibility limits: coolant flow rate 70-90L / min, temperature 21-25℃.

[0050] The safe operating range is the intersection of the above boundaries, with specific parameter values ​​as follows: Grinding wheel linear speed v_s: 25-40m / s (avoid the upper limit of 45m / s to leave a safety margin); Feed rate v_f: 0.4-1.4 mm / s; Cutting depth a_e: 0.005-0.03mm (single cut ≤0.04mm, cumulative cut ≤0.15mm); Dwell time: 0-1.5s (feed-in phase ≤ 1.5s, machining phase = 0s).

[0051] (III) Constructing a second-order response surface model Data is obtained through DOE sampling within a safe range to construct predictive models for grinding efficiency, accuracy, and surface quality.

[0052] 1. Using the "orthogonal Latin square + boundary reinforcement" sampling method, a total of 30 sets of parameters were selected (covering 90% of the safe range). Five gears were machined for each set of parameters, and the following data were collected: Grinding efficiency: number of parts processed per unit time (pieces / h), calculated as: 3600 / (single piece processing time t_total); Grinding accuracy: Tooth profile error f_fα (μm), cumulative tooth pitch error F_p (μm) (detected by Zeiss PRISMO in-machine measuring instrument); Surface quality: surface roughness Ra (μm), surface texture direction (detected by Taylor Hopson roughness meter); Process parameters: real-time power P(t), spindle temperature T(t) (sampled once every 10ms).

[0053] Some of the sampled data are shown in Table 1.

[0054] Table 1 Partial Sampling Data Serial Number v_s v_f a_e dwell efficiency f_fα Ra P_avg 1 30 0.8 0.020 0.5 8.2 4.8 0.82 16.5 2 35 1.2 0.030 1.0 10.5 5.5 0.95 17.8 3 28 0.6 0.015 0.3 6.1 3.9 0.75 12.3 With v_s, v_f, a_e, and dwell as independent variables (x1-x4), the least squares method is used to fit the model, which has the form: y=β0+Σβ i x i +Σβ ii x i 2 +Σβ ijx i x j For the model parameters and test indicators where i < j, they are as follows: Efficiency model (y1): β0 = 5.2, β1 = 0.15, β2 = 1.2, β3 = -2.8, β4 = -0.4, R 2 = 0.94, residual ≤ 4.2%.

[0055] Tooth profile error model (y2): β0 = 3.2, β1 = -0.06, β2 = 0.8, β3 = 35, β4 = 0.3, R 2 = 0.93, residual ≤ 5.1%.

[0056] Roughness model (y3): β0 = 0.65, β1 = 0.012, β2 = 0.06, β3 = 18, β4 = 0.04, R 2 = 0.91, residual ≤ 5.5%.

[0057] The model passes the R 2 test and the residual normality test, and is stored in the rsm_mode table (model_id = RSM001), associated with context_id = CTX2024051001.

[0058] 2. Model archiving Write the model fitting coefficients, cross-validation reports, and applicable context keys into the database to generate a model version fingerprint: F2E5D8C1..., ensuring traceability and reusability.

[0059] (IV) Hybrid algorithm optimization and target parameter selection Use the GA-PSO hybrid algorithm to solve the Pareto solution set and determine the target parameters in combination with decision rules.

[0060] 1. Parameter settings for the hybrid optimization algorithm Genetic algorithm (GA): population size 50, crossover probability 0.8, mutation probability 0.05, iteration 30 generations; Particle swarm optimization algorithm (PSO): number of particles 40, learning factors c1 = c2 = 2.0, inertia weight ω linearly decreasing from 0.7 to 0.9, iteration 20 generations; Optimization objectives: max(y1), min(y2), min(y3), constraint conditions: P_avg ≤ 18kW, T_spindle ≤ 85°C.

[0061] 2. Obtain the target process parameters Solving the Pareto solution set and target parameters yields 25 non-dominated solutions, through Figure 2 The "knee point identification algorithm" in the text selects the equilibrium solution, and the final target process parameter set is: v_s=32m / s, v_f=1.0mm / s, a_e=0.025mm, dwell=0.8s.

[0062] Corresponding predicted values: efficiency 9.3 pieces / h, f_fα=5.0μm (≤ design tolerance 6μm), Ra=0.88μm (≤ design requirement 1.0μm), P_avg=14.4kW (≤18kW); store in the pareto_archive table, associated with context_id=CTX2024051001.

[0063] II. Safe Method for Issuing Optimized Grinding Process Parameters This section describes the process of sending parameters to the CNC system based on some of the obtained target process parameters, and illustrates the process with reference to Figure 5. A parameter mapping quintuple is established between the parameters and CNC variables: "template parameter name—CNC variable path—unit scaling—read / write permission—verification rule." This mapping accurately maps key process quantities such as feed rate (mm / s), wheel speed (m / s), depth of cut (mm), and dwell time (s) to the domestic CNC system. A transaction mechanism is introduced when writing to the CNC system. Before writing, range verification, dimensional consistency, and mutual exclusion conflict determination are performed. After batch sending, immediate readback and consistency comparison are performed. If consistent, the process is committed; otherwise, it is rolled back and the safety bit is restored, while feasibility evidence and alarm information are recorded. Archiving and closed-loop updates are implemented, with templates, instances, model versions, optimization results, control instructions, and quality evidence uniformly archived. This archive drives incremental updates and retraining of templates and safety domains.

[0064] (a) Constructing a quintuple correspondence table As shown in Table 2, the correspondence table includes: Process parameter identifier, used to uniquely identify a gear grinding process parameter; The CNC variable path is used to specify the storage address or variable name of the gear grinding process parameters in the target CNC system; The unit scaling factor is used to define the conversion relationship between the units of the gear grinding process parameters and the units inside the CNC system. Read / write permission identifier, used to declare the read / write attributes of gear grinding process parameters in the CNC system; Parameter verification rules are used to verify the legality or logical consistency of parameter values ​​before writing the target gear grinding process parameters into the CNC system.

[0065] Table 2. Correspondence of Five-Tuples Process parameter identifier CNC variable path Unit scaling factor Read / write permission identifier Parameter validation rules feed rate CHPAR

[2011] 1000 (mm / s → μm / s) R / W 0.4≤v_f≤1.4mm / s; v_f≤(18000×0.88) / (2100×a_e×30) wheel_speed (grinding wheel linear speed) SYSVAR

[15] 1 (m / s → rotational speed conversion) R / W 25≤v_s≤40m / s; Rotational speed S=60v_s / (π×300)≤4000r / min (In this embodiment, S=2037r / min) depth (cutting depth) #500 (User macro variable) 1 (mm) R / W 0.005≤a_e≤0.03mm; Cumulative ≤0.15mm dwell (duration of stay) CHVAR[7] 1（s） R / W 0 ≤ well ≤ 1.5s; processing section well = 0 coolant_flow (coolant flow rate) IO

[302] .OUT 1 (L / min) R / W 70 ≤ Flow rate ≤ 90 L / min (80 L / min in this example) (ii) Parameter distribution 1. Pre-write validation Parameter range verification: Target parameters v_f=1.0mm / s and a_e=0.025mm are both within the safe range, passing the test. Mutual exclusion check: Confirmed that in "one-way trimming" mode, the two-way trimming parameter #501=0, and there is no conflict; Context consistency check: Verify that the current machine_id=GMC2000 and wheel_id=CBN300-120 match context_id=CTX2024051001, and pass; Expected Consistency Fingerprint Calculation: Apply the SHA-256 hash algorithm to the target parameter set to obtain the expected fingerprint A3B7D9F2... (64 bits).

[0066] 2. Batch distribution and readback comparison Batch delivery: Parameters are written sequentially via APIBatchWriteParam. After each parameter is written, wait 50ms to receive the ACK signal (total time 230ms). The process is stored in the transaction_log table (trans_id=TRANS001). Back-read comparison: Call BatchReadParam to read back parameters. The back-read set is consistent with the target parameters. Calculate the back-read fingerprint A3B7D9F2... and match it with the expected fingerprint.

[0067] 3. Commit / Rollback Mechanism Submission: Call Commit(TransID=TRANS001) to confirm that the parameters are effective. The system returns "OK", the status light turns green, and the transaction_log table records result=success. Exception handling: If feed=1100μm / s during readback (out of range), trigger rollback: call Rollback(TransID=TRANS001), follow the pre-write backup recovery parameters (feed=0.8mm / s), move to the safe position (X=0,Y=200,Z=100,B=0°,C=0), and record error_code=E102 (readback inconsistency).

[0068] 4. Process monitoring Monitoring parameters: Spindle power (threshold ≤ 14.4kW = 18 × 90%) and temperature (threshold ≤ 72.25℃ = 85 × 85%) are collected at 20ms intervals. Abnormal Handling: If the power suddenly increases to 21kW (exceeding the threshold), execute EmergencyStop()→Rollback()→MoveToSafePos() to trigger an audible and visual alarm, error_code=E111.

[0069] 5. Archiving and Closed-Loop Updating The "parameter distribution list, readback records, monitoring data, and processing results" are stored in the transaction_log and measurement tables. Based on the processing results (f_fα=4.9μm, Ra=0.86μm), the constraint range of a_e in the process_template table is updated, and the fitting coefficients in the rsm_model table are optimized to form a closed loop.

[0070] III. Grinding Process Parameter Safe Distribution System This system is based on Figure 1 The system architecture implementation, the functions of each module, the data interaction, and the correspondence with the attached diagrams are as follows: (I) System Architecture Database layer: Contains the template library (process_template), instance incremental library (process_instance_delta), context library (context_snapshot), measurement database (measurement), RSM model library (rsm_model), and Pareto archive (pareto_archive). All tables are categorized by... Figure 2 Content associations within.

[0071] Core module: Instantiation engine (execution) Figure 3 Instantiation process), optimization engine (execution) Figure 4 Optimized process), CNC mapping and write-back module (execution) Figure 5 The distribution process), visualization and auditing modules.

[0072] External interfaces: connect to domestic CNC systems, in-machine measurement equipment, IoT sensors, and user UI interfaces.

[0073] (II) Module Interaction Process Instantiation engine: Reads process_template and context_snapshot from the database layer, and then... Figure 3 The process generates a security domain and outputs it to the optimization engine.

[0074] Optimization Engine: Receives security domain data, and presses... Figure 4 The process completes DOE sampling, RSM modeling, GA-PSO solving, and outputs the target parameter set to the CNC mapping module.

[0075] CNC mapping module: Calls the parameter mapping table, according to... Figure 5 Once the process is completed, the transaction is sent out and the results are written back to the database.

[0076] Visualization module: Real-time display Figure 4 Pareto Frontier Figure 5 The system tracks the distribution status and monitoring curves, and generates audit reports.

[0077] (III) Key APIs and Functions Instantiation related functions: CreateTemplate(), SolveConstraints(), GenerateSafetyDomain().

[0078] Optimization related: DOESampling(), TrainRSMModel(), GA_PSO_Solve().

[0079] The following functions are involved in the transaction: BatchWriteParam(), BatchReadParam(), CommitTransaction(), and RollbackTransaction().

[0080] Audit-related functions: GenerateAuditReport(), QueryTransactionLog().

[0081] This specific implementation method uses a 20CrNiMo carburized and quenched gear with a module of 3mm and 50 teeth as the machining object. Based on a domestically produced CNC worm gear grinding machine and the Huashu HNC-848B system, it fully implements a multi-objective optimization and parameter safe distribution scheme for the gear grinding process: first, by collecting the rated parameters of the machine tool and the environmental conditions, according to... Figure 2 The data model is associated and stored in the database, and then followed Figure 3 The process involves building the table structure, defining the process template, and recording incremental instances. The safe range of grinding process parameters is determined through constraint consistency solving. Subsequently, DOE sampling is performed within the safe domain to construct second-order response surface models for efficiency, accuracy, and surface quality. The Pareto solution set is solved using a GA-PSO hybrid algorithm, and the target parameter set is selected. Then, a mapping relationship is established based on the five-tuple of "process parameter identifier - CNC variable path - unit scaling factor - read / write permission identifier - parameter verification rule," and the process is then... Figure 5 The transactional workflow completes pre-write validation, batch distribution, readback comparison, commit / rollback, and 10-50ms periodic online monitoring; finally, the entire workflow data is archived, driving closed-loop updates of templates, models, and security domains. The entire process relies on... Figure 1 The system architecture enables collaboration among modules, providing comprehensive coverage. Figures 1-5The core nodes, all parameters, formulas, and table structures were examined to verify the feasibility and security of the solution.

[0082] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A multi-objective optimization method based on a CNC gear grinding process database, characterized in that, Includes the following steps: S1. Obtain the rated machining parameters of the machine tool and the environmental conditions during gear grinding; S2. Based on environmental conditions, determine the capability boundaries of the computer-controlled gear grinding machine in terms of power, thermal load, and motion stroke, and refer to the rated machining parameters to determine the safe operating range of the gear grinding process parameters; S3. Within the safe operating range, obtain the three categories of measurement standards corresponding to each group of grinding process parameters: grinding efficiency, grinding accuracy, and surface quality, and construct a mathematical model to predict grinding efficiency, grinding accuracy, and surface quality. S4. Based on the mathematical model, obtain the target gear grinding process parameters corresponding to the target measurement standard, so as to guide the gear grinding process of the target machine tool.

2. The multi-objective optimization method based on a CNC gear grinding process database according to claim 1, characterized in that, Environmental conditions include the temperature range, humidity range, and coolant flow rate and temperature of the environment in which the gear grinding machine is located.

3. The multi-objective optimization method based on a CNC gear grinding process database according to claim 2, characterized in that, The safe operating range refers to the intersection of various gear grinding process parameter sets that respectively meet the capability boundaries in terms of power, thermal load, and motion stroke.

4. The multi-objective optimization method based on a CNC gear grinding process database according to claim 3, characterized in that, A mathematical model is a second-order response surface model that uses grinding process parameters as independent variables and grinding efficiency, grinding accuracy, and surface quality as dependent variables.

5. The multi-objective optimization method based on a CNC gear grinding process database according to claim 4, characterized in that, In S4, a hybrid optimization algorithm combining genetic algorithm and particle swarm optimization algorithm is used to generate multiple candidate grinding process parameter schemes, and the candidate grinding process parameter schemes that meet the target evaluation criteria are selected.

6. The multi-objective optimization method based on a CNC gear grinding process database according to claim 5, characterized in that, After obtaining the target gear grinding process parameters, they are sent to the CNC machine tool, including the following steps: Step 1: Establish a correspondence table between gear grinding process parameters and internal variables of the CNC system; Step 2: Based on the correspondence table, write the target gear grinding process parameters into the CNC system.

7. A multi-objective optimization method based on a CNC gear grinding process database according to claim 6, characterized in that, In step one, the corresponding relationship table includes: Process parameter identifier, used to uniquely identify a gear grinding process parameter; The CNC variable path is used to specify the storage address or variable name of the gear grinding process parameters in the target CNC system; The unit scaling factor is used to define the conversion relationship between the units of the gear grinding process parameters and the units inside the CNC system. Read / write permission identifier, used to declare the read / write attributes of gear grinding process parameters in the CNC system; Parameter verification rules are used to verify the legality or logical consistency of parameter values ​​before writing the target gear grinding process parameters into the CNC system.

8. A multi-objective optimization method based on a CNC gear grinding process database according to claim 7, characterized in that, In step two, before writing process parameters to the CNC system in batches, a desired consistency fingerprint is pre-calculated based on the target gear grinding process parameters. After the target gear grinding process parameters are written to the CNC system in batches, all written parameter values ​​are immediately read back to form a readback parameter set. Based on the readback parameter set, a readback consistency fingerprint is calculated. The readback consistency fingerprint is compared with the expected consistency fingerprint. If the two are consistent, the target gear grinding process parameters are confirmed to have been successfully issued and the parameter settings are submitted. If the two are inconsistent, a rollback operation is triggered to restore the CNC system parameters to their state before writing.

9. A multi-objective optimization method based on a CNC gear grinding process database according to claim 8, characterized in that, Both expected consistency fingerprints and readback consistency fingerprints are verification codes calculated by applying a cryptographic hash function to the complete parameter dataset. The complete parameter dataset corresponding to the expected consistency fingerprint includes the complete and final command data packet to be written into the CNC system. The complete parameter dataset corresponding to the readback consistency fingerprint includes the complete and actual parameter status data packet read back from the CNC system.

10. A multi-objective optimization integrated system based on a CNC gear grinding process database, characterized in that, This integrated system is used to implement a multi-objective optimization method based on a CNC gear grinding process database as described in any one of claims 1-9, including: The parameter mapping module is used to store and manage the correspondence table between the process parameters and the internal variables of the CNC system; The transaction control module is used to execute the transactional distribution steps, including compliance checks, batch writing, readback comparison, and result confirmation.

Images